Text

Testimony of Tj Adler on Behalf of Atty General for Commonwealth of Ma on Revised Sapl Contention 31,SAPL Contention 37 & Town of Hampton Revised Contention III to Rev 2 ETE Contentions.*
	ML20238F214
Person / Time
Site:	Seabrook
Issue date:	09/14/1987
From:	Adler T; MASSACHUSETTS, COMMONWEALTH OF
To:	;
Shared Package
ML20238F181	List: ML20238F179; ML20238F214; ML20238F223; ML20238F239; ML20238F262; ML20238F268;
References
	OL, NUDOCS 8709160028
	Download: ML20238F214 (166)
	v • d • e

_ __ __ . - _ _ _

UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION Before Administrative Judges:

Helen F. Hoyt, Chairperson Gustavo A. Linenberger, Jr.

Dr. Jerry Harbour l

\

) -

) i In +he Matter of ) -

) i PUBLIC SERVICE COMPANY OF NEW ) Docket Nos.

HAMPSHIRE, ET AL. ) 50-443-444-OL i (Seabrook Station, Units 1 and 2) ) (Off-site EP)

) September 14, 1987 1

TESTIMONY OF-THOMAS J. ADLER ON BEHALF OF THE ATTORNEY GENERAL FOR THE COMMONWEALTH OF MASSACHUSETTS ON REVISED SAPL CONTENTION NO. 31, SAPL CONTENTION NO. 37,

.AND TOWN OF HAMPTON REVISED CONTENTION III TO REVISION 2 (THE 'ETE CONTENTIONS") ,

i Department of the Attorney General Commonwealth of Massachusetts One Ashburton Place Boston, MA 02108-1698 (617) 727-2265 8709160028 870911 PDR ADOCK 05000443 T PDR

UNITED STATE 5 OF AMERICA NUCLEAR REGULATORY COMMISSION Before Administrative Judges:

Helen F. Hoyt, Chairperson Gustave A. Linenberger, Jr.

Dr. Jerry Harbour

)

In the Matter of ) .

)

PUBLIC SERVICE COMPANY OF NEW ) Docket Nos.

HAMPSHIRE, ET AL. ) 50-4'43-444-OL (Seabrook Station, Units 1 and 2) ) (Off-site EP)

) September 14, 1987

)

TESTIMONY OF THOMAS J. ADLER ON BEHALF OF THE ATTORNEY GENERAL FOR THE COMMONWEALTH OF MASSACHUSETTS ON REVISED SAPL CONTENTION NO. 31, SAPL CONTENTION NO. 37, AND TOWN OF HAMPTON REVISED CONTENTION III TO REVISION 2 (THE 'ETE CONTENTIONS")

Department of the Attorney General Commonwealth of Massachusetts One Ashburton Place l

Boston, MA 02108-1698 (617) 727-2265

UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION Bafora Administrative Judgas:

Helen F. Hoyt, Chairperson Gustava A. Linenbergar, Jr.

Dr. Jarry Harbour

)

In the Matter of ) *

)

PUBLIC SERVICE COMPANY OF NEW ) Docket Nos.

HAMPSHIRE, ET AL. ) 50-443-444-OL (Seabrook Station, Units 1 and 2) ) (Off-site EP)

) September 14, 1987

)

TESTIMONY OF THOMAS J. ADLER ON BEHALF OF THE ATTORNEY GENERAL FOR THE COMMONWEALTH OF MASSACHUSETTS ON REVISED SAPL CONTENTION NO. 31, SAPL CONTENTION NO. 37, AND TOWN OF HAMPTON REVISED CONT'ENTION III TO REVISION 2 (THE "ETE CONTENTIONS")*

I . __ IDENTIFICATION OF WITNESS Q. Would you please tell us your name and current occupation?

A. My name is Thomas J. Adler and I currantly serve as Prasident of Resource Systems Group and as Principal-in-Charge l

for transportation projects.

Q. I would like to begin this tastimony by questinr.ing you, Dr. Adler, regarding your background and the genaral natura of your work on this matter. What is your educational background?

To the axtant that this tastimony may address matters raised in other contentions filed by SAPL, it is baing sponsored by SAPL.

A. I raceived a Bachelor of Science in Civil and Environmental Engineering from Cornall Unversity in 1972, a Master of Science from Massachusetts Institute of Technology in 1975 and a Ph.D. from M.I.T. in Transportation Systems in 1976. While completing my doctoral dissertation I was a fellow at the Joint Canter for Urban Studies of M.I.T. and Harvard.

Q. What has been your professional exparianca sinca

! complating your formal educational training?

l A. Over the period 1976 to 1986 I was a professor in the Rasource Policy Canter, an acadamic and research program of the Thayer School of Engineering at Dartmouth College. At Dartmouth I taught graduate and undergraduate-leval coursas in the araas of transportation systems analysis, transportation engineering, transportation planning, computer / mathematical modeling techniques, statistical analysis, and computer science. I diracted a program of research for clients such as tha U.S. Departmant of Transportation, U.S. Departmant of Energy and others in tha genaral area of computer modaling of transportation systems. I also consulted on a part-time basis for major transportation consulting firms such as Cambridga Systematics (Cambridge, MA) and Gannett Fleming Transportation Engineers (Harrisburg, PA).

Q. What was the nature of the rasearch activities pursuad by you and your colleagues at Dartmouth's Rasource Policy l

Canter?

A. My cantral focus at the Rasourca Policy Canter, and the focus of our academic program, was on the scianca of

computer modaling of complex system. We ware very much concarnad with formalizing proceduras to guide the development and application of large computer models. Th9 Center organized sevaral major conferences on avaluation of computer models of enargy supply and demand, and I personally diracted raviaws, for federal agencies, of several large transportation models.

Q. Have you also been involved in tha actual development and application of transportation-ralatad modals?

A. Yes. For axampla, at Dartmouth I directed a project for the U.S. Department of Transportation whose product was a simulation model of transportation energy use. I also directed anothar projact which involved development of a traffic network simulation modal. I have also been haavily involved in the davalopment of techniques for collecting data to be used as input to transportation models. I was co-author of a manual distributed by the Urban Mass Transportation Administration which describes naw tachniques for using bahavioral intentions data in transportation models.

Q. For h>w long have you servad in your current position as President of Rasource Systems Group?

A. I have been President since Octobar 1986, whan Resource Systems Group was formally incorporated.

Q. What is the nature of your current work at Rasourca Systams Group?

A. Besides my administrative responsibilities, I have maintainad research and project activities similar to those that I pursuad whila a profassor at Dartmouth. For axample, I

am under contract with the' Ohio Department of Transportation to independently review.a'larga transportation network modeling project. I am also directing a project to estimate input ,

i parameters for an urban transportation modal for Tampa, Florida, under subcontract to the transportation planning firm of Parsons Brinckerhoff Quade & Douglas. I have been involved with ona of our other principals in review and enhancements of an unclassified model deval.opad for use by the Office of the Joint Chiefs of Staff in the Department of Dafense. And I hava also worked for numerous municipalities and private concerns in ,

projects involving traffic impact analysis. I have served as Principal-in-Charge at Resource Systems Group for the project I will be describing in this testimony, our Seabrook Evacuation Time Estimation project.

Q. Besides you, who are the other principals at Resource Systems Group and what ara their backgrounds?

A. Dr. Dennis Meadows is one of the other two founders and principals of Rasource Systems Group. Dr. Meadows is also Icofessor of Engineering and Diractor of the Resource Policy Center at Dartmouth College. Ha is an internationally l

recognized expert in the field of computer m3deling and ]

l simulation, and ha is the author of several books and numarous l journal articles documenting and evaluating applications of computar modaling of complex systems. He has twice served as l Senior Scientist at tha International Instituta for Appliad Systems Analysis in Laxanbourg, Austria, most recently as ,

l Program Director. Dr. Meadows also servas as exacutiva 1

1

-4 -

l

t diractor of.INRIC (International Natwork of Resource Information Centers), a confederation of over twanty international institutes involvad in computar modaling and resourca analysis. Prior to joining tha facdlty at Dartmouth he was a professor at M.I.T. His activities at Resource Systems Group hava included development of simulation models for management training, and ha has served as an advisor on our Seabrook Evacuation Tima Estimation project.

Dr. colin High, the other principal of Resource .

Systems Group, is a profassor in the Dartmouth Environmental ,

Studips Program. He has formal academic training and extansiva axperianca in tha fields of geography and environmental science. He also has directad several major projects which involvad tha davalopment and application of computer simulation models. At Rasource Systems Group, Dr. High has directed resource modeling projacts for international clients and serves as an advisor on projacts involving ramote sensing. He has taught coursas in interpretation of aerial photography and directed a NASA-sponsorad projact in that field. He was directly involved in the ir.terpratation of aerial photographs of the Seabrook area seacoast which wera compiled for the Seabrook Evacuation Tima Estimation project.

More datailad resumas for all three of the Resourca Systams Group principals are attachad to this testimony (Attachment 1).

Q. Are thara other professional staff at Resourca Systems Group who have contributed to the work on your Seabrook Evacuation Tima Estimation project?

l i

.A. Yas. We currently.have seven full-tima professional-staff members 1who'workLon modeling and transportation-ralatad

't

.projacts. Most of thase' individuals were involved:in some part u

of Tthe Seabrook Evacuation Time Estimation project'. Staphen f 1 L !Hastings, a Senior Project Associata with a Master of Scienca- 1

. degree from'Dartmouth in System Simulation,_ conducted most of

.the computer analyses with'the assistanca of Jeffrey Sidell, !

who also has his dagree from Dartmouth in computer Scienca.

Two other Project Associatas, Leslia Rimmer and Megan Haney, '

assisted-in several data analysis tasks. All of thase tasks i

were performed under my direct supervision.. During my

~

testimony I wil1 rafer to these Resource Systems Group staff and principals as "we*. TheLwork11s familiar to ma in that, while I did not perform all of tha work personally, I did personally specify and supervise that work.

l II. OVERVIEW OF THE TESTIMONY Q. Would you please describe briefly tha scopa of the work that Rasourca Systems Group performed for the Massachusetts Department of the Attorney General?

A. We were asked generally to review and avaluata tha validity and reliability of tha Seabrook Station Evacuation Tima Estimates and the adequacy of the Traffic Managament Plan Updata'as incorporated into NHRERP Volume 6.

O. How did Resource Systems Group approach this work?

A. We divided the work into three ganeral tasks:

1) Evaluate the accuracy / reliability of tha assumptions and data that wera used as inputs to the analysis reported in Volume 6;

- 6-

2) Datermine the effacts on Evacuation Time Estimates (ETEs) of those inputs which.

we found to be inaccurate or otherwise unraliable and, whare possibla, insert tha better assumption /more accurate data; and

3) Assass the importanca of factors that were not explicitly considered in the Volume 6 ETEs and determine their effect on the ETEs.

In complating parts of tha first task we relied in part on data and analysas performed by other consultants ratained by the Massachusetts Attorney General. Specific -

I contributions will be detailed later in this testimony. In several very important areas, it appeared that the assumptions and inputs used in Volume 6 were not appropriate and so I wanted to tast tha effects of alternative assumptions / inputs on the ETEs. In ordar to allow direct comparisons of these rasults with the ones raported in Voluma 6, the best procedura would be to' employ the identical computer modeling protocol used by KLD in preparing thosa ETE estimates. It was also important to independently verify the KLD results, as is commonly done in othat computer modeling applications. Thus, we concluded that we naeded to have direct access to KLD's computer model. On July 8, 1987, KLD provided us with a compiled version of the computer program and computer tapes of the input files it used for the ETE study documented in Volume

6. Ovar the intervening pariod we hava completed over 75 separate model runs and sensitivity tests using I-DYNEV, the computer modal developed by KLD and used by KLD in preparing - -

ETEs for tha NHRERP, Ravision 2. While the compiled I-DYNEV i

f

program allowed us to complate several simulation runs, it did not. allow us to verify the implementation of tha methods described in Volume 6 as we would have been abla to do had we been providad tha computer source code. Although we had requasted I-DYNEV's source code, KLD refused this request.

On reviawing Volume 6, I also idantifiad several important factors that were not explicitly considared in the KLD analyses, and we performed additional analysas, to avaluate and quantify tha effects of those factors. -

This testimony will detail the wer< that we completed and the results that were obtainad which are germane to this proceeding.

Q. In the work that you did, did you apply mathods and procedures which are ganerally acceptad in you profession?

A. Yas.

Q. Bafore you go into furthar detail describing your work, would you please first summarize the most important results?

A. Thera are several findings which individually are quite important, but the most important single conclusion daals J

with tha overall reliability of the ETEs prasented in Volume 1

6. In our view those ETEs are unreliable, espacially for the summer scenarios. The results of our study indicate that the overall tima to avacuata tha antire EPZ on a summer waekend whan tha beachas ara crowded is not 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months and 15 minutes as KLD has astimatad, but is at least 11 hours1.273148e-4 days 0.00306 hours 1.818783e-5 weeks 4.1855e-6 months and 15 minutes. Wa also found that the vehiclas in tha beach araas are the ones which expariance the longest ETEs and that, because of the extremaly limited numbar of roads leading off the baaches, cars will back up in such long queuas thara that many beach-goers will ba unabla to gat off the baach strip (Rt.1A) for 8 to 10 hours1.157407e-4 days 0.00278 hours 1.653439e-5 weeks 3.805e-6 months , depending on whera thay parked their cars.

The principal author of the KLD study has assertad that the accuracy of the ETEs for a sumn.er weekand, as reported in Voluma 6 of the NHRERp, is plus or minus 10 percant. In fact, we found, using the same computer model employed by -

Volume 6's authors, KLD Associates, that tha Voluma 6 ETEs consistently and significantly underestimate actual ETEs under both plausibla and likaly evacuation conditions.

One reason why KLD has seriously underestimated ETE's is that it made a saries of assumptions about aspects of any avacuation which ara unlikely to prevail in raal-lifa conditions. Two such assumptions are (1) that all evacuating l vahiclas will take tha routas out of the EPZ which the plans prafar; and (2) that all traffic control posts ("TCPs") will ba staffed, with barriers and traffic cones in placa, at tha moment an evacuation begins. But avan if the sumi.ter weekend evacuation ware to procaed exactly as planned, which we do not balieve would happen, we found that ETEs will still be significantly longar because some of the critical inputs appear to be in arror. In particular, as the tastimony I presented jointly with Drs. Bafort and High has shown, tha number of vehicles which can occupy tha baach areas on peak summer waekands is at least 50% greatar than reportad in Volume 6.

e Thus, KLD's astimates of evacuation times for its summer weekand scanarios (i.e., Scanarios 1 and 2) reflect not the timas to evacuate beaches which ara at 100% of capacity, but the times to evacuate beachas which are only about 2/3 full.

\

In addition, tha KLD analysis ignores altogether (i.a., does not aven factor into tha modal) certain important real-life alements such as the additional traffic created by thousands of drivers who will be returning home to pick up othar family mambers. -

Together, the effect of just the few factors I have specifically mentioned abova is to increase the ETE by more than 96% over KLD s astimate--from 6:15 to 12:15--for the i

summer weakend evacuation of tha antire EPZ. Note that I say -'

more than. In doing our model runs, some of the affacts that lengthen avacuation times were not modelad for all locations in the EPZ. For exampla, we evaluated tha affects of returning commutars only at two locations, raprasanting only a small fraction of tha total of such vehicles. Had wa modeled this affect for all locations, ETEs would have bean shown to ba evan longer than reportad hera. We did not do this additional l modeling bacause the objactive of our work was to determine tha overall accuracy and reliability of the Volume 6 ETEs, not to prapara a full set of " alternative ETEs." As I describa later in my tastimony, my axpert opinion is that the complata set of ETEc for Saabrook Station's EPZ can and should ba re-dona in a mannar that producas a much more reliabla set of ETEs than those in Volume 6. .

While most of our analyses assumad full evacuation of the EPZ (Region 1), we did evaluate the effect on ETEs for a baach-area-only evacuation (Region 10) for the summar weekend scenario (Scanario 1), with the effects described above includad. In this analysis wa also assumed a 50% "snadow" evacuation of the ramainder of Region 1, following the recommendation about this factor described in tha testimony of Drs. Zeigler, Johnson, and Cole. The results are shown in Figura 1 (next page) in comparison to the values shown in Volume 6, Table 10-9 (p. 10-11). The times shown, in all cases, rapresent tha timas to evacuata the baach araas only, i.e., to get all cars off of Rt.lA or the raspective beach roads. As this figura shows, we estimated beach evacuation timas approximately twice as long as reported in Voluma 6. The time it takes to evacuata to 2, 5, or 10 miles away from tha baaches is somawhat longar.

We also found several important areas whare omissions and ambiguities in Transportation Managamant Plan elements of

-Volume 6 will rasult in an ovarall reduction in the efficiency of tha evacuation, comparad to what would ba possible/ desirable. For axampla, tua management of emergency and commuters' vehicles raturning into the EPZ has not been carefully considered, with the result of likely extandad delays for all of these vehicles at several Access Control Points (ACPs) and Traffic Control Points (TCPs).

Q. When you mantion the tarms "Scanario" and "Ragion" what are you referring to?

nwa s 1

Time Required to Evacuate Beach Areas j NH Plan vs. New Estimates :

i

^ f Evacuation Shadow o m Beach ;

E

' Closing e

.. y d Hampton g ,

Seabrook o

6'.

Salisbury m

,

2 n 4 Plum Island a t

1 Evacuation Zon 3 NH Plan Volume 6 Estimate

@ New Estimate NW, /

i

$$N kk , /

.e ' 'c e'J:, /

',' ' :,' gg's '

$>: 1 7 , 't . ,

Sqh

t - {

"j ; : c Wgf General

. , J ', ;f' [ .

? Evacuation

, $ 7, ' C E,lfR;' O % 8 '

,,e$ ,, ',, 0,e\'; Hampton j

' ,"', n " .

f '

" >c <

's g (=4= a, ,

,', , "'}el' > TSeabrook i

u. : , . z > c, v,wer vnw \ g Evacuation Zone l QdL,,u,s ' * , ;'s , 'v YY \ - .

'g'4l,;;

l f,;; g^u

Salisbury U.

~

!'Q/;if '",' ' a :g:p '

n y ,l; ,'

e ' ':,"",',;Y)h,,, \ %

J .

g 3 ,

[( ' . ' 'T <uC n,

Plum Island b v ':": ^ ,

s

' ' ' ' , ' s J1 Q lg%., *,

& \ ',-

g,,'?

t Sources: Resource Systems Group '

VS.

\

NHRERP Volume 6 l

l

- 11a -

A. In Volume 6, ETEs were calculatad for different conditions depending on saason, day, time of day and weather.

A given combination of these factors constitutes a " Scenario".

A table from Volume 6 listing "Scanarios" used is attachad to this tastimony (Attachment 11). " Regions" are those subparts of the EPZ for which ETEs were calculated for each "Scanario".

A table listing the "Ragions" used in Voluma 6 is attached to this tastimony (Attachment 12).

Q. Figure 1 is titlad, "Tima Requirad to Evacuata Beach Araas." Would you axplain further what this figure shows?

A. Figure 1 shows the amount of time required to claar tha immediate beach area only under two evacuation conditions:

beach closing (uppar part of figure) and a general evacuation of the entire Emergancy Planning Zone (lowar part of figure).

The amount of time required for vehicles to claar out of the baach araa is longer in the general evacuation condition because more cars enter the network in front of tha evacuating baach traffic. For each evacuation condition, two times are presantad for each of the beach areas: the results raported in Voluma 6, and tha onas that we computed with tha assumptions 1

that I dascribed above. Note that all of tha timas shown rapresant the amount of time requirad to claar only the roads immediately adjacent to the beach. For Hampton, Seabrook and Salisbury, thay generally represant the amount of tima it would take to clear all evacuating vehiclas off Rt.lA.

Thare are two other important points of clarification ragarding thesa time astimatas. First, all of the Voluma 6

evacuation time estimates contain a systamatic bias due to an inappropriate interpolation method that underrepresants 2

evacuation times by up to 25 minutas (this will be discussed latar in this testimony). The "New Estimates" reported in Figure 1 are calculated using a method that eliminates this bias. Second, all tha timas presantad in Figure 1 (and in I Voluma 6) actually reprasent tha time that would elapse after an assumed ganeral evacuation message that, itself, is assumed ,

to occur 25 minutes after the beach closing message. So, for -

axample, tha 5:40 reported in Volume 6 as tha evacuation time for the Hampton beach area actually repres9nts 6:05 after the beach closing messaga. The use of this general avacuation order as the base time for comparison is, however, mislaading when applied to the beach closing-only condition since a general evacuation order is naver issued. We have reported in Figure 1 beach closing evacuation timas as they were calculatad in Volume 6. So, to determine the amount of time that baach avacueos will actually be "on the road," 25 minutas should be addad to all of the times shown in Figure 1 (and in Volume 6).

Thus, for example, if only the beaches areas are ordered to be ,

cleared, we found that it will take 8:40 to do so for Hampton Baach and 5:40 for Seabrook Baach on days when the beachas are full. ,

Q. You have shown baach avacuation timas of over 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months } -

for Hampton's beach araa and 10 hours1.157407e-4 days 0.00278 hours 1.653439e-5 weeks 3.805e-6 months for Salisbury's; yet, in a sansa those beaches " evacuate" even on the busiest days in much less time than this. How do you axplain the differenca?

A. Thera are several important differences between the amptying of the baach araa on a busy summer day and the circumstances of an emergency evacuation. First, as raported in the testimony of Dr. Albert Luloff, data from a recent beach ;

survey by AEL Associates show that on a " normal" day vehiclas leave the beach area at a relatively uniform rate over a 5 hour5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months period from 1:00 PM to 6:00 PM. However, in an emergency evacuation, everyone would be attempting to leave the area i 1

almost simultaneously. Second, over 20% of tha individuals on the beach have come by a means other than driving a car (most s by walking) and presumably will raturn to their accommodations also without driving a car. Almost 30% of the beach visitors ,tf are staying in tha area more than 3 nights. Neither of these L J

populations would have cars on the road during a " normal" afternoon period when tha beach emptias. By contrast, all 3 4

would be part of the traffic stream in the event of an emergancy evacuation. In addition, all year-round rasidents and araa workers would be: leaving at the sama time. Finally, .

there will be additional <individjtals:in the areas between the beachesandthemajornorth-southhigbwayswhowillba evacuating and who would'normally not'be on the highways at

/

that time. In summary, the major diffarenca between tha 4

amptying of the beach area on a busy summer day and th9 conditions of an emercaricy evacuation is a substantial increase in the number of vedic 19s leaving the araa and a substantial <. i 1

l raduction in tha time interval over which thay will try to antar the road networ'k. .)

o

.) \

l j

l

__________-______D

4 Q. Could you dascriba how an avacuation trip from the Hampton Beach State Park would likaly procaad in the ganeral evacuation scanario on a busy summmer day?

A. Assuming everything proceadad according to the Voluma

\

6 Plan for a Ragion I/Scanario I avacuation, a beach closing massage would first be issued and all baach visitors would ba instructed to laava the beach area. Thosa who were abla to reach thair cars tha most quickly would likaly be the first to exit the lot. If they were to follow thair dasignated avacuation route, they would turn right (north) onto Rt.lA and continua to Rt.51 wast. For tha first few vehiclas, this would be a ralatively quick journey. However, those who had to .

"round up" a family of several childran would be blocked from

niexiting the parking lot by a substantial quaue of the 1,500 p

q iL other v$hicles attempting to exit this lot onto Rt.lA. Others

J' e

. sho parkad at locations to the north of the State Park lot would also craate a substantial queue to the north on Rt.1A.

Only 25 minu,tes after the baach closing, a general avacuation would be ordared and additional cars would fill tha roadways to g.

l tha north'and west of the State Park. According to the data provided in Voluma J, it would take 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months and 5 minutos (from

/

tha tiac of the'bacch closing massaga) for tha last cars out of

.tha Hampton Beach Stata Park lot to maka tha 1 and 1/2 mila l'

journa31, north to Rt.51~. This translatas to an averaga forward prograss of approximately 1 and 1/2 car langths par minuta.

}. ;

According to the new anc.irata dascribad abova, this same trip vi +

) would take 9 hours1.041667e-4 days 0.0025 hours 1.488095e-5 weeks 3.4245e-6 months and 50 minutas, raprasenting forward 1 ,-

i

prograss of lass than one car length par minuta. In fact, prograss would not be continuous, as soma vahiclas would wait for long periods of tima with no movamant at all, and than mova ahaad in spurts of several car lengths.

Figure 2 (naxt paga) shows tha number of vahicles which would remain along Rt.lA in Hampton ovar the pariod following a baach closing announcement (and subsaquent ganaral evacuation). According to the I-DYNEV runs which ara reportad in Voluma 6, ovar 3,000 vahicles (ovar 7,000 paopla) will still be stuck along Rt.1A in Hampton's baach araa 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months after tha baach closing. According to our new estimata, approximately that sama numbar would ramain aftar 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months and over 18,000 people would ba in tha quaue along Rt.lA after 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months .

Tha langth of tha delays and tha axtant of congestion would rapresant traval conditions wall outsida tha realm of any of the evacueas' prior axpariance. Data from tha AEL Associates survey, raportad in tha tastimony of Dr. Albert Luloff, indicate that substantial numbers of evacueas will simply abandon their cars if little forward prograss is made over an axtendad pariod of tima. All of tha I-DYNEV runs reported in this testimony assume that averyone in tha evacuation zone is abla to finally evacuata. Howevar, thera ara inherant bahavioral uncartaintias in how avacuaes collectively and individually will raspond to thesa conditions. I am unabla to stata that full avacuation is likaly or avan possibla without substantial intervention bayond what is dascribad in Volume 6, undar tha conditions that would axist.

Figure 2 g :

n

,n --

.[ '

x '

)

Vehicles in the Hampton Beach Area Ib '

'During a General Evacuation v (Summer Weekend Mid-day)

-r p

),

!2 ,.L ,

_a- _

y Yy '

a 120xi A. -

' I o

., E NHPlanEstimates h .{ ,

10000' New Estimates eA 1 g. ... > >

3 2 E, gg.

7 - '[' : n o, s x 4

) l I

6000 -

e' '

M4 s

y = ,

j" i, -

, . j Yf 40CO - :

[

s

! J_

27V) - /.

g 0-- ru 1 . b -

'/

{ -

9 .0 1

.L 3 4 5 .6 7 8 9 10 s

V f .

.+

, .;:." Time From Bzach Closine;(Hours) o

,i ,

y v ,.

44 6

.n 1

's

?

Sources: ittsocce Sytten.s Group vs.

"Ni&ERY Volume 6 t .\

f .i ' - 16a -

-g r -.

E j A i

..w_m : __ _ m - _ . .

_____-m_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ . _ . . _ _ _ _ _ _ _ . _ _ . . _ _ . _ _ _ _ _ . . _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ . _ _ _ _ _ _ _ _ _ . . -

0 'Would you explain'your last statement further? l A. Yas, cartainly. Given avacuation spea'ds of less than f

l. one car length per minute for periods'of time up to 8 or 9 hours1.041667e-4 days 0.0025 hours 1.488095e-5 weeks 3.4245e-6 months (just to get out of the Hampton Baach State Park parking lot, travel up Rt.1A for a mile and a half, and get on Rt.51 west), I am uncertain that a vehicular evacuation will work. I just do not know whether paople will stay with their cars that i long, traveling much slower than most of tham can walk. The evidenca I have, from the baach survey by AEL Associates, indicates that many will indeea abandon thair cars. .Tha problem, of course,_is that not everyone can walk out. There 6 are many beach-goers with infants and small children, as well as many who are elderly, infirm, handicapped or simply not fit enough to walk a faw miles. Will those who abandon their cars laava their cars in positions which block those who remain with thair cars? I simply do not know. But it is such a real possibility, particularly for a place like the Hampton Beach State Park lot, from which the nuclear plant is clearly in view, that prudant planners naed to plan additional layers of l

assistance to address this circumstance. I am particularly concerned about the State Park lot for tha following reason:

many of tha cars in that lot are at the end of the evacuation queue for the entire south baach araa in Hampton, those cars will sit there in the lot, not moving at all, for hours before the rest of Hampton's beach area clears out enough to allow the cars in that lot to begin amptying onto Rt.lA. During the hours before the lot begins to empty, many cars may well be

abandoned, and this has the real potantial of creating

" gridlock" ineide the lot unless thera is substantial additional planned intervention. The NHRERP Traffic Managament Plan calls for only a single traffic guide to ba stationed at l

tha point where tha Hampton Beach State Park lot empties onto Rt.lA, and that guida's job is described as follows:

"[e]ncouraga all traffic to move north along Routa lA." (See Vol. 6,'p. I-1). More than a single emergancy worker certainly naeds to be stationed at tha Hampton Beach State Park to deal with the potential for " gridlock" there, and emergency strategies naed to be devised to address this egress problam, as well as to address the neads of hundreds, if not thousands, of baach-goers who may abandon thair cars and attempt to walk out.

Q. You stated that your analyses are based on tha same computer modal used in the preparation of Volume 6. Would you axplain this statement further?

A. A major issue in this case is the reliability of the numbers and assumptions that serva as inputs to the ETE calculation. I was retained by the Massachusetts Attorney General's office to evaluate the effects of these numbers and assumptions without changing the basic mode]4ng methodology employed in Voluma 6. The reliability of the computer model used to calculata tha ETEs, tha I-DYNEV system, is a separate issue that is addressad in some detail by Dr. Ceder in his testimony but which I did not pursua in detail for the purposes of my work. I will make some general commants on I-DYNEV later I

h in my testimony, but the major focus of my work was on testing 1 1

l tha effects of alternative inputs to tha modal rather than on critiquing tha model itself.

To be more specific, wa obtained from KLD a copy of the I-DYNEV model system which was verified by KLD staff to ba functionally identical to tha. version that thay used in praparing Volume'6. We also obtained a full set of tha inputs used by KLD for all of the results reportad in Volume 6 as well as inputs used in KLD's further sensitivity tests. In all cases that I will describe here, we used this varsion of I-DYNEV without modification and bagan with the exact inputs used for Volume 6, modifying thesa inputs only as appropriate for each individual computer run.

III. ANALYSIS OF THE CONTENTIONS Q. Have you read the various contentions filed by the interveners regarding ETE issues, specific 411y Revised SAPL Contention 31, Town of Hampton Revisad Contention III and SAPL 37?

A. Yes. Copies of thesa contentions are attached to this testimony (Attachment 2).

Q. Would you please detail the findings of your work as thay relate to these contentions?

A. Certainly. Since some of tha important issues are l raised in more than one of the contentions that you listed, I will organize my discussion around issue areas and cite tha relevant contentions / bases as I procaed. I will start with a 19 -

central issue with respact to the ETES, tha KLD estimata of beach (" transient") population in the Seabrook seacoast area.

[ For the purpose of computing ETEs, KLD was primarily interested l

in computing the vehicle population in the beach areas. To do this, they relied on several saries of aerial photographs from which the beach area's " parking capacity" was astimated.

According to volume 6, pp. E-4 to E-5, this " capacity" includes the stalls in all marked parking lots, curb space along roads, othar opan accessible lots, private driveways, front yards and ,

back yards. KLD's interpretation of tha resulting estimate is given on p. E-5 of Volume 6:

We believe that these capacity estimates represent a reasonabla upper bound to the number of possible parked vahicles in the indicated areas.

Town of Hampton TOH III/ Basis A and SAPL 31/ Basis 20 coutend that KLD had inadequate data on actual parking utilization during truly peak periods to accurately infer the araa's " parking capacity."

To assass this contention, the interveners first obtained the photographs used in KLD's analyses and requested additional detail on KLD's methods so that the Volume 6 1 estimates could be replicated. Unfortunately, KLD discarded (

l tha datailed worksheets on which they recorded their counts from aach slide. Having these worksheets would have allowad us l

l' to conduct a more complete audit of their work. While actual car counts from the KLD August 1985 photos were replicatad approximately (as described in tha tastimony of William I

Befort), the " parking capacity" astimate could not be j j

I replicated from'the: protocols recounted-in Mr. Lieberman's' July .

2, 1987, deposition.,

i Two aerial overf1+1ghts of the.Seabrook EPZ beach areas-i were. conducted by. William Befort in July 1987 to provide additional photographic data for our work. Interpretation of that photographic ~ material gave two important' sets of data b- points for our ETE work: an-estimate of the actual number of-vehicles present in the beach areas'on July 5, 1987, and an estimate-of " parking capacity" as defined in Volume-6. We used the latter estimate, in a way consiste c with'the methodology described in Volume 6, to. recalculate (using I-DYNEV) ETEs for the summer 1 weekend evacuat!Jn scenario. -For this'I-DYNEV run, int used all of KLD's inputs except those which reflected the beach area's vehicle capacity, where~we substituted our new data. -The results are shown in Figure 4 (see the bar labeled

" Beach Parking Capacity"). Even though the total number of evacuating trips for the entire EPZ is only 15% higher than the estimate used in Volume 6, the ETE, increases by 51% (from 6:15 to 9:25 for Scenario 1 full EPZ) just by changing this single input fact 6r, because the beach population disproportionately.

, loads the most difficult bottlenecks in the roadway network.

Q. Who actually prepared the vehicle population and

- " parking capacity" estimates from the July 1987 aerial

- photographs?

A. Drs. Colin High and William Befort were directly responsible for that work, though I was involved collaboratively in developing the protocols used and in l

l

_ 21 -

c__=_-- -- . _ _ _ -

reviewing the work. There has been separate testimony filed which outlines the procedures and results of this work.

Q. Dr. Adler, before we move on to other issues contained

! in the contentions, were there other data and analyses prepared by other consultants to the Massachusetts Attorney General which you used in your work?

A. Yes. Social Data Analysts conducted a telephone survey of EPZ residents which provided demographic data and information on likely evacuation behavior. This survey responds directly to the concern raised in SAPL 31/ Basis 6, which asserts that the survey relied upon in the KLD study has a large non-response bias. The Social Data Analysts survey also collected important information that goes beyond the scope of the survey relied upon by KLD. Dr. Stephen Cole has submitted testimony detailing the methodology employed in this survey. Resource Systems Group was provided with tabulations of the survey responses, a data tape containing the individual computer-coded responses, and the original written survey forms.

AEL Associates prepared resident population and employment estimates and projections for the New Hampshire EPZ towns. They also conducted a beach survey which provides demographic information and indications of likely evacuation behavior. Dr. Albert Luloff's testimony details these data and accompanying analyses.

I also worked closely with Dr. Avishai Ceder, who has been reviewing some of the traffic capacity, traffic flow, and engineering aspects of Volume 6 and of the I-DYNEV modeling system.

i:

'Q. . Did.you do any work related to.the-contentions dealing

'with the effects of roadway capacity on ETEs?

-A.. Yes. 'Both TOH III/ Bases B'and C and SAPL'31/ Bases 7, i.

9,;15 and 18 deal with.the issue of how capacities for the area's' roadways were estimated in Volume 6. In fact, those

' bases. cover'many different-issues related to roadway capacity which we evaluated.

Q.. Is traffic capacity an important factor in determining-ETEs?

A . .. Yes, definitely. ' Generally, the evacuation time is a direct function'of the capacity of the constraining roadway sections. A simple example illustrates this point. It would-

.take 10 hours1.157407e-4 days 0.00278 hours 1.653439e-5 weeks 3.805e-6 months for 10,000 vehicles to evacuate through a roadway

~

section which has a capacity of 1,000 vehicles / hour (10,000 divided by 1,000 = 10). .If the capacity of'this roadway section were 25% lower-(750 vehicles / hour), it would take 13 and!1/3 hours.to evacuate the same 10,000 vehicles (10,000

' divided'by 750 = 13.33), or 33% longer. So, roadway capacity has a direct effect on ETEs.

Q. Would you summarize the description contained in Volume 6 of the procedure used by KLD in preparing the capacity estimates in Volume 6?.

(

A. As described in Volume 6, p. 1-10, KLD conducted a-

' survey of the " entire highway system within the EPZ and for some distance outside." KLD claims this survey detailed the numerous characteristics of each highway section which are necessary to estimate traffic flow capacity. KLD then

. - _ - . _ _ = _ _ - _ - -- - .

7_

n; ,

)

.. h

': estimated maximum flow capacities for-each of four' general roadway. types and, s'eparately, for freeways and freeway ramps, based.on procedures outlined in the Highway Capacity Manual

( 1985). In addition, "such estimates were obtained empirically at representative. intersections.throughout the EPZ" (p. 3-3). .

l These. capacities are. decreased'by I-DYNEV.by 15% "when the !

. traffic is' moving under congested conditions" (p. 3-9) to reflect the: detrimental effectsoof driver behavior under'those circumstances. KLD.also says that capacities at intersections' -

were adjusted to reflect the effects of evacuating traffic flow (if.any) from: competing approaches. 'The capacities were further reduced in scenarios which assume inclement weather:

~

by 20% for rain and by 25% for snow.-

Q. What did you do to verify and review the work related to' traffic. capacities performed by KLD?

A. I-used a three-step review process. . First, I wanted' to, verify the survey data that KLD had' collected on the area's roadway-characteristics and KLD's empirical observations of traffic flow. However, KLD had discarded all original records of1their roadway survey and had nct maintained formal reccrds of their traffic flow observations. Rather than replicate the full survey, I personally visited the area and. inspected several key intersections to determine whether they were accurately represented in the Traffic Control-Post (TCP) and Access Control, Post (ACP) diagrams (see Vol. 6, Appendices I and L) and whether they were properly represented in the I-DYNEV model. On one occasion'Dr. Ceder and I toured the area

, together. ,As I will describe below, we found some-important

+

discrepancies.-

, The second step of the review-process involved l-checking I-DYNEV to verify that'the capacities specified were properly reflected in the model's outputs. We again found discrepancies which I will detail.

Finally, I worked with Dr. Ceder to determine whether the capacities assumed by-KLD are, as they assert,

'" conservative estimates" (Volume 6, p. 3-3, emphasis in

- (- . original), and, generally, whether those. capacities are reasonable for the conditions that they are intended to represent. Dr. Ceder will describe his work on this issue in his testimony.

Q. Would you now please detail your findings related to roadway capacity estimates?

A. We found overall that the actual achievable capacities of the evacuation' roadway network are lower than the capacities represented in Volume 6, for five primary reasons:

(1) The' actual road network has gaometric restrictions not recognized /modeled by KLD; (2) The I-DYNEV model. appears to apply incorrectly the 15% reduction factor'for congested flow conditions on critical freeway on-ramps; (3) The effects at intersections of considerable traffic flows caused by commuters returning home is.not modeled; (4) The effects of dicabled vehicles have not been explicitly considered; and (5) The effects of inclement weather have not been adequately represented.

j

The first-listed problem is best illustrated by reference to one of the critical bottlenecks that appears in the summer weekend evacuation scenario, the intersecti9n Rt.110 and I-95.' 'This intersection is critical'to the beach area evacuation because Rt.110 serves simultaneously as: (1) the primary evacuation route for a large number of beach evacuees from Salisbury, (2) a planned " overflow" route for Seabrook beach (see p. 1-31), and (3) the primary return route for commuters who work in areas southwest of the EPZ and who are returning to both-New Hampshire and Massachusetts towns within the EPZ.

The Traffic Control Post Diagram (Vol. 6, p. I-19)

(Attachment 3 hereto) for this intersection fails to show in the southwest quadrant a wide, grassy median with raised 6 inch curbs that connects the two triangular islands shown. A modified version of this TCP, which shows this impediment, is attached to this testimony (Attachment 4). As routed, in order to access I-95 southbound from the left turn off Rt.110 westbound, evacuating vehicles would have to travel over this median. In fact, these vehicles are being routed onto an off-ramp which is posted with large "Do Not Enter" and " Wrong Way" signs. Vehicles exiting I-95 along this ramp are being only " discouraged" (not prohibited) from making this movement, which may result in head-on collisions. The I-DYNEV model assumes that the capacity of this wrong-way use of an off-ramp, involving traversal of a raised, grassy median is the same as for a normal, paved freeway on-ramp. This assumption grossly

L.

R - overestimates the achievable capacity under these conditions.

-It is a reasonable assumption that a passenger car'will slow to 5 mph to climb the curb and that approximately a 25 foot gap will be left.between cars to allow for lurching after the curb is negotiated. In these~ circumstances, the capacity will be t-lowered to 660 vehicles / hour, almost half (56%) of the capacity assumed by KLD.

A run of I-DYNEV was made in which.the capacity of this ramp was reduced to 660 vehicles / hour. All other KLD inputs remained unchanged. The result was an evacuation time 0

of 8:50, a 41% increase in the overall ETE. (Region 1,

3. . Scenario 1).

Importantly, the above analysis does not consider the fact that there are many camper-trailers (over 400 in one

' Salisbury lot), low-clearance sports cars, and other vehicles 1

that simply may not be able to traverse the median.

Undoubtedly, some will try to do so, and these vehicles would further impede traffic flow. Nor does the Volume 6 analysis consider the effects of rainfall or snow storage on the ability ;

i of vehicles to cross the median.

Q. The second capacity-related problem that you identified deals with the I-DYNEV modeling of congested flow on ramps. Would you please describe the work you did related to I l

this problem? -1

)

A. Yes. Volume 6, p. 3-11, says that freeway ramps were assumed to have a capacity of 1170 vehicles / hour and that the capacity is reduced to 85% of this value under congested flow 1

_ _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ _ _ . . _ _ _ . _ s

L h f conditions tol account for congestion effects. The resulting capacity should be 994 vehicles / hour. We.were provided by KLD-

-with their actual computer output of the run which represents.

the summer weekend evacuation (Scenario 1). This output

, contains many' details not reported in Volume 6. In checking

-through this output, I discovered that the actual simulated 1

discharge rate through I-95 ramps which serve as bottlenecks to the evacuation ranged up to 1096 vehicles / hour during periods.

'f. congested (Service' Level F) flow.

o This represents.10%-

. higher capacity.than'what is stated was used in* Volume 6.

Since this overstatement of capacity occurs at a network bottleneck, the result is a proportional understatement of ETEs.

Q. The third capacity-related problem that you listed earlier deals with the effects of commuters returning home.

Would you please describe the work that-you did related to this problem?

A. The KLD analysis, as reported in Volume 6, includes directly only the trips of commuters who work in the EPZ but live outside. That analysis considers the effects of other commuters returning home only in a very limited way--by assuming that the capacity of two-way road sections will be reduced to a level corresponding to a 90%-10% split of evacuating vs. returning traffic. The trips of commuters who both live and work within the EPZ and who must travel along with evacuating traffic are not considered at all. However, the most significant effects of the returning traffic are (1) at' intersections, where returning traffic streams will in many l

J [. _n o,

cases'directly conflict with evacuating traffic, with resulting

.significant declines in capacity.for evacuating flows and (2) along evacuation routes where commuters' vehicles are attempting to travel in~the same direction as evacuating traffic.

Our analysis proceeded first by evaluating the general magnitude of this problem, then developing case examples to illustrate the effects on ETEs. Volume 6 states (at p. 4 ,9 )

that 95% of the commuting population would leave work within 30-ininutes of the evacuation order. This would create an immediate surge of traffic that would be far greater than any experienced under " normal 1 circumstances." Data from the Social Data Analysts survey shows that, under normal conditions, only one-third of the EPZ's-resident workers leave work in the afternoon peak one-hour period (4:00 to 5:00 p.m.). The remainder of the EPZ's, resident workers leave work during other times of day, generally between 3 and 6 p.m. See Figure 3 (next page). Thus, the " normal" rush hour reflects only a small percentage of the commuter traffic activity that would occur under the weekday evacuation scenario. Therefore, contrary to KLD's assumption, in a real evacuation it would take considerably longer than normal for the commuting population to return home.

Some of the adverse effects of " return home from work" trips on the evacuating traffic flow were represented in the Volume 6 study by using an assumption of 10% counterblow I traffic in the calculation of roadway capacities, as described

Figure.3 j I

Work Departure Time Distribution 1

Seabrook Station Emergency Planning Zone )

40 -

i

. 35 - l l

1 gg 30 -

!Ei: -

g3 25 -

g ,.9 r5 20 -

E

.$a h 15 -

n -

10 -

@]

$~5, 5- :?

- ' hi 0" E " """ -

Noon 1 2 3 4 5 6 7 8 9 10 11 Midnight Time l

i l

Sources: Resource Systems Group, SocialData Analysts vs.

NHRERP Volume 6 _ 29g _

g ,

W h ,

i m /

<inzVol'me16 u at p._3-8'.- However, these'" return home fromework"

' trip's were..not explicitly modeled incany of the evacuation scenarios,.norLwere they modeled as affecting intersection i capacity,_even thoughLthey would add a significant number;of trips'toLthefroad' network during an evacuation. Furthers the-assumption of only 10% counterblow traffic ldoes.not'appearoto-L "be realisticfbecause it reflects a serious undercount of return trips.: . Table 5-3 of Volume 6-(p. 5-7): indicates that 1986 Lemployment wichin the EPZ.was 68,084 and that 31,298 of-those-employees reside outside the EPZ. ~ Since'all-evacuating trips by. permanent' residents are modeled as originating from the

^ place 1of' permanent residence, KLD assumed that many of these

.36,786 commuters; return _'home prior to evacuating. Using the average vehicle occupancy of l'.16 for employees, as provided in Table 5-3 of Volume-6 (p. 5-7),.a total number of 31,712' trips home could be generated by people returning home from work within the EPZ.

Information in Section 5 of Volume 6 indicates that 26,420 EPZ. residents' commute to workplaces outside the'EPZ.

Using the 1.16 employees per vehicle figure mentioned above, the potential exists for 22,776 trips from places of work outside the EPZ to places of residence within the-EPZ. Thus, in total,.up to 54,488 commuter trips home could occur that have not been explicitly modeled in any of the Volume 6 evacuation scenarios. These trips would generate additional

. crossflow, counterblow, and evacuation flow. In the cases of scenarios 3 and 4, mid-week scenarios when employment is at

l l100%, 54,488 commuter trips home would represent a 53% increase in the total number of-trips that KLD modeled as occurring during an' evacuation. Most of these. commuter trips would-begin-l .over.a'short period'of' time at the onset oflthe evacuation, causing a heavy flow of commuter traffic that would cause increased. congestion, impeding both.the evacuating'and commuting traffic.

To' illustrate the. effects of commuter traffic, we added only 200 additional commuters.to the " combined run"

. renctted earlier and the effect was an increase in ETEs of 10 minutes. Since thi's represents only a small percentage of the 54,000 potential commuter trips, the effect of explicitly including all. commuter trips would be substantial.

Q. You also listed a potential capacity-related problem with disabled vehicles. Would you explain this issue further?

A. Volume 6 discounts the problem of disabled vehicles by asserting (1) that disabling accidents are unlikely in an evacuation due to low travel speeds and (2) that disabled vehicles could, in any case, "be pushed onto the shoulder" (pp. 12-3, 12-4). The Lieberman affidavit for summary disposition of SAPL 31 (March 25, 1987) contends further that disabled vehicles on bridges or other choke points would either be. pushed out of the way or would be stored in a location where traffic could move around-the disabled vehicle, "even if it means encroaching, somewhat, into the incoming lane of travel on a two-lane road." In this latter situation, it is asserted that " traffic flow in the inbound direction would take

g ,

p-advantage'of'any gaps in-the evacuation, traffic."

.Nowhere does b VolumeM6_. indicate-how many;of the' evacuation' routes in the EPZ p

p have. adequate shoulders to store disabled vehicles, w

. Presumably,-this'information was gathered.in the field survey referenced in Volume 6-(p.-1-10), but records from the' logs that were compiled by KLD staff.in surveying these routes were discarded by KLD (eTen though these records are explicitly referenced in Volume 6). During winter conditions, snow banks frequently arellarge enough to prevent disabled vehicles from ;

1 being pushed from the traveled way onto the shou'1 der. The-assumption that' incoming vehicles will find gaps in evacuation traffic that is diverting around such obstructions is unsupported at best given the heavy volumes of both evacuating-and'of counterblow traffic. Further comments on this' issue appear in Dr. Ceder's testimony.

In response to the contention that vehicles removed to the shoulder will reduce effective capacity, the Lieberman i affidavit compares the capacity of 1728 veh/hr/ lane used in the KLD study for the area's freeways to observed flows of 1800 during construction operations when lane widths were reduced (asLcited in the Highway Capacity Manual). The comparison cited, however, is not valid because it does not include the effect of. additional capacity reductions from use by non-commuting populations present in large numbers on a weekend evacuation. The Highway capacity Manual recommends additional reductions between 10 and 25% to account for lower levels of familiarity with the road network for weekend and recreational

' traffic. If the 1800 veh/hr/ lane' observed capacity figure is adjusted as recommended to account for this effect, itLreduces to between 1350:and 1620, substantially lower than the capacity of 1728 assumed in the ETE study.

Neither the KLD study in Volume 6 nor the Lieberman affidavit' indicates how long it is estimated for a tow truck to respond to an' incident and return to its assigned locatio'n and, therefore, how many incidents could be' managed oy each: truck.

The'ETE study. assumes that low vehicle (average) speeds will .

l result in relativelyL few accidents which produce disabled vehicles. However, Volume 6 assumes'an average speed for incoming' buses of.40.-50 mph. Undoubtedly, many commuters will be rushing home to pack up their families at speeds reaching 40-50 mph between intersections. Thus the potential for high-speed accidents exists. Even though the' inbound bus

" average" speeds will be higher than speeds the buses will commonly use inside the EPZ (because most of the travel outside the EPZ will be on interstate or other high-speed roads) the j travel within the EPZ will have to average about 40 mph to maintain the cited overall 50 mph average. If evacuating vehicles are forced into the opposing lane by vehicles with i

mechanical or other failures, the potential for an accident with these high speed counterblow vehicles is significantly l

l increased.

There are at least three locations where evacuating vehicles are routed into lanes normally used by vehicles traveling in the opposite direction: the intersections 1

depicted in TCPs B-AM-06 (p. I-19), B-SA-06 (p. I-28) and D-HA-02 (p. I-39). At.all of these locations, a significant

. potential exists for head-on conflicts between re' turning commuters and. evacuating vehicles. .Overall, the likelihood of disabling accidents' appears to be greater under evacuation conditions than in normal circumstances. .See the testimony of U Dr.Eceder for additional testimony on this point.

I have one final point regarding disabled vehicles.

The beach survey conducted by AEL. Associates found that 13% of those-surveyed would abandon their car if-little forward progress were made in 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months and over 1/3 of the beach-goers 1

would abandon their car if little progress were made in 3 l hours. Since even.the evacuation time estimates in Volume 6 show longer waits than these for1certain beach evacuees, prudent evacuation planners must consider the likelihood of significant vehicle abandonment.

Q. Finally, you mentioned the effects of weather on roadway capacity. Would you explain your findings on this q

)

issue?

A. TOH Contention III Basis B argues that the effects of adverse weather' conditions on roadway capacity and, correspondingly, on ETEs are not adequately accounted for in the ETE study. The ETE study states that " estimates of roadway capacity must be determined with great care" (p. 3-1). It is I further stated that the calculated effects of weather on highway capacity were " based on limited empirical data" (p.

l 3-1). In fact, as argued in this contention, no site specific I

r -

' data on the extent of delay' caused by these road hazards were collected.' In addition,.the'effecte of' fog were simply i

, . surmised to txe equivalent to the effects of rain. The effects of rain were in' turn estimated to be equivalent to those

' observed in studies' conducted elsewhere.

.The'Lieberman-affidavit of March 25, 1987 (TOH III),

states that "public' officials indicated that ocean fog'is an

. unusual occurrence during the summer months" which " generally

~

dissipates at 9 a.m. or may appear after sunset."(p. 6)'No names of these o'fficials.are'given. Other data, however, indicate that ocean fog is common in the summer months.

.Further, ra'in and fog-together are not uncommon. Weather

-records for Portland, Boston, and for Pease Air Force Base (NH) provided by:the Applicant.in response to discovery in fact support the contention. Those records show that " heavy fog" is experienced between Stand 6; days per month throughout the

- period.of May to October in Portland. Over that same period general fog was observedHat Pease-(1957-1979) on 46% to 54% of the days and 18% to 22% of the total hours. This record indicates's fog presence much greater than " unusual".

In Volume 6 and in the Lieberman affidavit, the discussion of effects of fog asserts that a sudden rain scenario, as modeled with a 20% capacity reduction, is responsive to this issue. .The only empirical evidence cited is from the 1985' Highway Capacity Manual, which states that "10 to 20% reductions (in capacity) are typical and higher percentages are quite possible" (emphasis added). No data specifically on

effects.ofifog are cited, nor was any data on the icfluence of fog on trafficfflow along the New Hampshire seacoast 1

collected. :Given the prevalence of this condition, prudent- ]l

. planners would not fail to obtain empirica11 observations of the L . effects of fog on the capacity of critical evacuation links in the Seabrook EPZ. Such empirical observations could be made quite easily. Further, prudent evacuation planners would not fail to make. empirical observations of the. combined rain / fog condition, given how easy such observations are to make and how critical this capacity-reduction factor is.

In his affidavit.of March 25, 1987 (SAPL 31),

.Lieberman also states than no estimate of the effect of roadway ice on capacity was found in a literature review. He then -

contends that "[i]n the absence of such data, 25% reduction in capacity as used for snow, appears acceptable." (Lieberman affidavit, pp. 19-20). Lieberman supports his poaition by stating (p. 20) that, although "[h]ighway capacity during.an ice storm may be less than'that during a snowfall", this is offset because trip generation times will be less given that driveways would not need shoveling. However, the latter effect will be small compared to the capacity reduction service because the time at which trips are added to that network is not.as important as the number of trips. Moreover, Lieberman states that "under severe. ice conditions, in the absence of sanding, some highway sections with extended upgrades may become virtually impassable." (Lieberman affidavit, March 25, 1987 (TOH III p. 7)). To jastify ignoring this condition,

s. o

Lieberman states: "It'is my understanding that sanding ~is a-prevalent procedure.in.this area." Id. However, in his.,

deposition, Lieberman. states (June'29, 1987, p.

. 145):-

O.. . Do.you have any idea how many sanding vehicles.would be needed?.

.A.- No, because.I don't believe that sanding operations should take place during an evacuation. They're more of an. impediment than a help.

Q. Why is that?

A. Because they're very slow-moving, sluggish vehicles, and they'would slow.down traffic. For the same reason I wouldn't want to see snowLplows out th'ere during an evacuation. i In any case, sanding and plowing operations would be severely impeded by traffic jams, barriers and cones, causing sand truck cycle times to be substantially longer than under normal conditions and resulting in ice accumulations that would

- make, upgrades " impassable." This will be inevitable unless additional specific provisions are made in the plan for providing extra sand trucks, beyond the current available fleet, to offset increased cycle times.

Q. You mentioned earlier that the ETE computations provided by KLD are biased on the low side by as much as 25 minutes because of a conceptual error in the interpolation

< method they used. Would you please explain this further?

A. The I-DYNEV outputs used by KLD to compute ETEs show the cumulative number of vehicles that have passed through a given area (e.g., the edge of the EpZ) at half-hour intervals.

Since all of the vehicles may finally clear out of an area in

. q s

between two 30-minute intervals, an adjustment procedure was, used to determine when, during the last 30-minute interval, the

~

area actually' cleared. The method used by KLD to determine the .

ETEs to the nearest 5 minutes, as reported in Volume 6, is described in the Lieberman deposition of June' 29, 1987 (p. 117):

^

We generally used as a matter.of course

' half-hour outputs. And-then by interpolation, sensible interpolation, we were able to

-pinpoint.what the ETE is, certainly within plus or minus ten minutes and probably within plus or minus five minutes. .

This " sensible interpolation" was accomplished by the following method (based on Lieberman deposition pp. 120-121):

30 x (Vehicles 2 divided by Vehicles 1)

Where: Vehicles 2_is the change in total i vehicles discharged over the last 30-minute--interval in which vehicles-remain in-the selected area; Vehicles 1 is the change in the number of vehicles that. passed through a selected area in the half-hour period proceeding that in which the area emptied out; and 30-is the length of the interval, in minutes.

This method assumes implicitly the.t all of the roads within this area empty at an even relative rate throughout the final 30-minute interval in the evacuation. In fact, it may be that traffic remains on only a single road during this last period and that it takes much of that period to empty that one road. Thus, the correct way to perform this interpolation is to compute the quantity shown above for each roadway in the selected area separately. Since the ETE is defined as the time at which an area is cleared of all vehicles, the ETE adjustment is correctly computed as the maximum of these values.

l.

A simple example may be helpful in illustrating this point. Assume that there are six roads leading out of the area being evacuated, each of which carries 500 vehicles when loaded to capacity during the next-to-the-last 30-minute interval in the evacuation. During that period a total of 3,000 cars would leave this area (= 6 x 500). Assume, then, that all but one of these roads empties completely during that next-to-the-last period, and that the final road has 400 vehicles left to clear. Since that road can carry 500 vehicles in 30 minutes,

, it would take an additional 24 minutes (= 30 x-400 divided by 500) to clear this final road. The KLD interpolation method would compute, instead, the ETE final period adjustment to be only 4 minutes (= 30 x 400 divided by 3,000). Thus the KLD interpolation would, in this example, underestimate the actual evacuation time by 20 minutes as a result of this incorrect methodology. More generally, the methodology used by KLD introduces a systematic bias that underrepresents actual ETEs by between 0 and 30 minutes.

Q. You referred earlier to work that Dr. Luloff conducted to estimate and project resident population and employment levels. Would you first explain which contentions / bases you used this material to address?

A. TOH III/ Basis E and SAPL 31/ Basis 3 both reference the concern that the New Hampshire seacoast area's "significant !

growth" in population and employment will quickly make the ETE estimates obsolete and that the effects of this growth should be anticipated in the current NHRERP. The December 1986 RAC i

a

,,e

)

review (" General Comments," #7) recommended that:

If additional significant', growth is -

anticipated, then ideally the impact of future increases in employment.on evacuation time ,

estimates should be considered. .(Similarly,

if growth in the population of permanent !

" residents is anticipated then that growth should also be considered in determining

, ' evacuation times.)

1

-There is a very good reason for requiring that the effectn'hf'this growth be considered in this Plan now, even

, .I though the growth will presumably be reflected, as it is i /.

realized,.in future plan updates. Crowth will result in longer, ETEs unless appropriate highway improvements yre put into place

,/

to offset the larger vehicle loadingss- The pianning process l

required to achieve those improvements is off en a lengthy one, ranging over ten years for major < projects involving fedeial funding and Manvironmental impact studies. ror those '

j improv'ements to be in place in time for then to avert unacceptable inOreases in evacuation times, growth must be anticipated several years in advance.

t Q. What did you do to estimate the likely effects of future growth? /

, i

{ A. Dr. O;loff projected resident population levels for L the New Hampshire EPZ towns, as described in his pre-filed <

testimony and in various tables and written material that he transmitted to Resource Systems Group. We translated his New f Hampshire population estimates into I-DYNEV: vehicle loadings using the methodology described in Volume 6. All other KLD inputs, $nciuding the 1985 Massachusetts EPZ tovn populations

\

reported'in Volume 6,, remained unchanged. The resulting f 1

40 - l 1

- a J

0:

,lf ' lp' j,p((4 <

{g,n ; ffl gem '

.0; y o

M o

s. .

I-DYNEV runs show anD00, increase in ETEs for Scenario 1 (the

!V w6cle EPZ)-for.the yearfi990 po]pulation' estimate',,an 11%'

o@ 1, j

l-

?

. [ 'V p 'y p,'

g m vy+ ep ,s e

}"g[dincrease'.for indIeasefor2010.

j 1995, Also, a because 2/4 'increa,se the

~

for the year 2000, a Town of Seabrook,is F

T J' -( ,

3 j,

growing-rel'atively, faster than the other towns.in the area, the

& network bottlenecks are shifteh by this growth.

M '

'Q. Is there additione)lkwork of this type that.should be ifF a.6 .,

y~ et f ,c .clomplSted in order to , asses s the f ull 'ef f ects on ETEs of ' f uture

<i .

growth?' ,

- A. : Yes. Growth in employment, which has been even more rapid than:' population growth!(Volume 6, p. 5-1), was not .

factored <directly into ourtanalyses. Similarly, the effects of a

.additionalretail'facibties, hotels /motelsandother estatliishmehid ' constructed since the 1981 NRC (Kaltman) study on which the Tolume 6 estimates are based, were not considered and properly-should be. .The effects of future increases in beach area pardtag should also be considered. A comparison of the August 11,"1985, aerial photographs relied upon in Volume 6 q

.j' .\

with current c6nditions indicates the formation of new parking lots in the Interim-2 years. This is a trend which will likely y continue as population pressures on the beaches escalate and )

e- which sho,uld shd ,

'f would be ar;icipated by prudent planners.

8

/.

~ Finally, our analyses did not consider the effects of any k anticipated future growth in the Massachusetts towns whose traffic mixes bith New Hampshire vehicles in many key i evacuation loca.tions. All of these growth effects, taken together, will lengthen ETEs well beyond the levels indicated

['byour'rmore limited analysis.

l 1 i 6 ../

j u___. l(l

Q. Were there.other I-DYNEV runs that you did which showed significant differences in ETEs from those reported in Volume 67 A. Yes. We conducted numerous other analyses, including one set which was intended to test the sensitivity of the ETEs to input values which were derived from the telephone survey commissioned by 'KLD. One particularly important input KLD

,obtained from this survey is the assumed number of residents who will evacuate in each car. SAPL Contention 31/ Basis 6 argues that the KLD survey is potentially biased to non-response and that there is "therefore no reasonable basis for assuming that the . . . estimates of average person occupancy of vehicles evacuating the EPZ or other data derived !

from the survey are accurate." Basis 19 specifically asserts that "[t]he estimate of 2,6 people per vehicle for permanent residents is unrealistic." This value is important because it directly determines the number of r.esident vehicles which will be loaded onto the roadways during the evacuation. KLD used a value of 2.6, person / car, which is based in part on the KLD survey results and in part on somewhat arbitrary assumptions regarding likely household behavior (Vol. 6, p. 2-5).

The Social Data Analysts survey included a specific question regarding the number of cars that would likely be used in a general evacuation scenario. Tabulation of those data J yields 2.3 person / car. This value is more reliable than the KLD input because it is derived from a survey with lower non-response bias and it does not depend on arbitrary

.o,a

.S .I ,k '

W'

, ,, n 3.']; .3 >.

3 -

!y . , e . s n

y '

7 33 7 -

.i

',I - ii

~h/, J srs$ptionsb'1the;.analys@abogt~ iikerly household behavior.

<3

,Thei}testimonytofDr,htephen?colec6scribesin.greaterdetail i

s , : ..

V the-. problems Mir.h,thglsurtrey reliei upon by KL0. When we n

3 !

1 conducted ar.;i.-DYuBV run 031ng 2.3l person dcar, and held all

. i U v otbtr irl nuts constant at MLD's valu49, .it resulted in an

~

s .

h, overall STF increaE9 of 'li for Scenario 1s y

l k 0: Od.St otner' types of ETE estimatec did you prepare S o .

^

? utH.ng I-JYNEV?). i A u .

t A l ', ' 'Ne perf ormed a det of; analys ts that tested the .

a

.\, I s

assumptions c5 cat,tt.y way the overall traffic management plan would work. Volum'c-6Jcomputes ETEs based on a traffic g .

a. E canagement pla'n that is.-Jescribed in various sdctiona of that s

A 4i [ document, including sections 9, 12, and appendices I, J, T. and L. In particular, the evacuation routings,are designated it. j appendices J and K while the other mentioned sections specify

- how inte?cections would be controlled to facilitate traffic flow along those routes.' There are many critical assumptions made in those sections of Volume 6, assumpti' ens which are i

contesyedinvariousadmitttdcontentions, inc'uding SAPL l

cntention 31/ Basis 4 ("[it in unrealistic] tp assume that traffic management and control measures are in.e&Eect at the time the evacuation is ordered") aM Town of Hampten Contention III/ Bases (C)2 and F.

Q. What specific tests of this traffic management plan did you conduct?

., A. We evaluated the effect on ETEs,0f three different l

typea of likely traffic manageraent proolems:

. 1:. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

m j Tf ,

i "Vl - > '

I M '

'(1)1 . Delayed staffing-of'some ofLthe traffic a, . control posts (TCPs);

L (2) : Evacuees choice of= host locations and-'

L routes'other-than those assumed in the

' Volume'6 analyses; and

!? .

(3)- . Non-optimal' traffic flow control at intersections.

Q. Would you please describe the ways in which you

represented these effects and the'results'that you obtained?-

A. SAPL'31/ Basis 4 asserts that the ETEs presented in b' 'Volumr 6 are not realistic.because they fail to account for

'delayad staffing of TCPs and ACPs. The analyses reported in

, Volume 6' assume that all transportation and access' control-l-

, . posts are fully operational at the time of-the beach closing J message. 'See Volume 6, p. 10-70. There are at least three clocations'at which TCPs are configured to use lanes in ways

, !substantially different from the ways they are used.in normalL flow ~ conditions. The I-95/Rt.51/Rt.101C intersection is E . changed by placement;of 29 traffic cones to allow westbound evacuating' Rt.51 traffic- to travel in the eastbound lane (TCP

.D-HA-02, p. I-39). (This TCP diagram is Attachment 5 to this testimony.). This operation depends, in turn, on the prior

-placement of traffic control measures (see TCP F-EX-04, p.

I-74) (Attachment 6) on Rt.51 in Exeter to stop any flow of

eastbound traffic from continuing to the I-95 interchange.

.This-eastbound movement is to be physically prevented by

' placement of a " Truck" at the Exeter'TCP. The net effect of this arrangement is to enable use of both Rt.51 overpass lanes (over I-95) for evacuating traffic, doubling the evacuation o.,

,1 ,< capacity?overg what would"otherwise be available from the single lE w'e'stbound. lane. Similarly, at the intersection of Rts.1, lA andfl10'(TCP B-SA-06, p. I-28) (Attachment 7), an eastbound lane-of Rt.110 is converted through the intersection into a westbound evacuation lane. In this case, there are no

' downstream controls specified to eliminate encroachment by eastbound. vehicles. originating from the numerous curb cuts along Rt.110 west of this intersection. Finally, an I-95 southbound off-ramp is converted to an on-ramp to double evacuating capacity at the I-95/Rt.110 interchange (TCP

~B-AMN-06 p. I-19) (Attachment 4).

As noted above, the I-DYNEV results presented in Volume'6 assume that all of the traffic control posts are fully operational from the beginning of the beach closing (which is assumed to occur 25 minutes before the general evacuation). We ran the I-DYNEV Scenario 1 model assuming, instead, that posts

[ at the Rt.110/I-95.and I-95/Rt.51/Rt.101C intersections were not fully operational until 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months and 45 minutes into the general evacuation (or approximately 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months after the beach closing). The result was a 7% increase in evacuation times over those reported in Volume 6. Longer delays in staffing these posts will have proportionally larger effects on the ETEs.

Q. Is it reasonable to assume that at the b I-95/Rt.51/Rt.101C intersection, the TCP will not be fully operational for two hours after a beach-closing announcement?

A. Well, as I understand the situation, the Town of Hampton is not cooperating in evacuation planning, and all of

1 i

its TCP's, including this one, are to be staffed by members of the New Hampshire State Police. The most recent Summary of {

1 Personnel ReSCJrce Assessment for the NHRERP (Aug. 1987) indicates v., p. 37 that 97 state police are needed for NHRERP l duties from a total of 185 available state-wide. The State Police do not have the cones and barricades they need to set up this TCP, however. These are supposed to be delivered by a truck driven by personnel from the New Hampshire Department of Transporta. tion. Thus, these personnel must first report to duty, load the truck with cones and barricades, then drive to each TCP, unload the designated cones and barricades, then drive to the next TCP, and so on. Also, as I understand it, each town's TCPs are to be staffed and equipped in an order of priority established from a priority list provided by the police chiefs of each town. For Hampton, the highest priority I

TCPs are the ones on the beach roads. Therefore, given that I the state police for this TCP may be coming from some distance, .

1 that the early arrivals will be assigned to the beach areas, and that the TCP can't be established until the cones and barricades arrive, it seems unreasonable to assume that this TCP would be operational in less than 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months .

Q. The second traffic management problem that you raised deals with the host destinations and routes that evacuees would use. Would you describe what you did to evaluate this problem? !

A. The Volume 6 analyses assume that all evacuees will travel to their designated host communities. This assumption does not appear reasonable for the large number of beach l

transients from the Hampton beach areas, who are assumed to travel to the City of Manchester, New Hampshire, as a host location. Under summer conditions, a large percentage of these evacuees come from areas south of the Hampton beach area.

According to the plan, all Hampton beach transients are to be i routed to the west and north out of the beach area toward q i

Manchester. Many are routed west along Rt.51 and Rt.101C. ;

i We did additional I-DYNEV analyses to determine the H effect of Hampton beach visitors evacuating to locations other than Manchester. In one set of runs, we assumed alternatively i

that 1/3 and 2/3 of the Hampton beach area visitors would evacuate directly to the south along Rt.lA, eventually connecting to I-95. The effects on ETEs were 15% and 47%

increases (over the Volume 6 general evacuation times for a ,

summer weekend midday evacuation), respectively, for the 1/3 and 2/3 routings. Even if Hampton beach evacuees were successfully routed to the west along Rt.51, it is likely that, at the I-95 interchange, many would choose to merge onto the I-95 on-ramp rather than continuing west along Rt.51. In my opinion, i't' is unrealistic to assume that those evacuees who i are from Massachusetts and who would have proceeded laboriously along Rt.51 will continue along this route to an unfamiliar destination to the west when they have an opportunity to access I-95 either northbound or southbound. The travel lanes of I-95 at.this interchange will be moving at 65 mph in uncontested flow in both directions. We tested the effect of 2/3 of the Rt.51 traffic accessing I-95, rather than continuing west l

l

along Rt.51. The result, as estimated by I-DYNEV, is an increase in ETEs for Scenario 1 of 15%, holding all other KLD inputs constant.

One additional set of I-DYNEV runs that we performed tested the sensitivity of the ETEs to. assumptions about the routing of evacuees, assuming that they in fact choose to travel to their assigned host communities. The analyses presented in Volume 6 assume that evacuees are able to determine the fastest route to their intended destinations (p. -

B-1). Unfortunately, the shortest route under evacuation conditions may not be the same as the shortest route under normally congested conditions. It is not likely that individuals from outside the area will know the fastest route to their destination, as Volume 6 assumes, especially when that route is along unfamiliar roads.

Both residents and beach visitors in the Portsmouth and Hampton, New Hampshire beach areas are shown in Volume 6 using a route to the Dover / Manchester host sites via Rts.236, 16A and 16 through Maine in a route to Manchester via US 4 in New Hampshire. These routes are circuitous and use roads'which are less well known than the direct routes to Manchester and Dover, Route 101 and the Spaulding Turnpike. We ran I-DYNEV for two conditions: the first assuming that these New Hampshire beach evacuees would not use the circuitous evacuation routes through Maine and New Hampshire, and the second assuming, additionally, that Portsmouth/ Rye-area residents would not use these circuitous routes. The runs

showed 7% and 11% increases, respectively, in ETEs for the

-summer weekend Scenario 1, holding all other inputs constant.

Q. Would you describe the work that you did to evaluate the third traffic management problem that you listed earlier, "non-optimal traffic flow control at intersections"?

A. Many of the evacuation routes merge at various points with other routes. At the intersections where two or more major routes merge or cross, traffic controls are specified in I-DYNEV to indicate the allocation of time between the competing directions. These controls are analogous to traffic lights; they provide alternating opportunities for the traffic streams at each of the approaches to move through the intersection. The I-DYNEV runs used in preparing Volume 6 assume that these intersections are controlled with 75 seconds between successive " green" indications for a given. approach

(" cycle length"). In addition, each intersection approach is assigned a " green time" ranging between 10 and 65 seconds, selected to optimize the flow through the intersection.

Since many of the intersections do not currently have ;

traffic lights and since timings for existing traffic lights f; I

are different from those used in I-DYNEV, the Volume 6 analysis l implicitly assumes that the assigned traffic guides will manage l

this allocation process. Thet'e are two problems with this, i

i d

First, the traffic guides are not instructed as to the proper timing for the intersections that they are assigned to ;

I control. Through much of the evacuation, traffic at the major {

{

intersections will be queued in lines that extend farther than i

l

___--__-_____ -- 1

the guides;can see, and.there will likely-be a tendency ~ simply-to allocate time evenly among all approaches. Even if the-Eguides. knew the optimal timing for each evacuation condition,.

they will~not necessa'rily have stop watches to gauge the l ; timing. The.second problem with this assumption of optimal signal timing is that many of the intersections, for which the I-DYNEV Volume 6 runs assume this type of signal control, are not designated as TCPs or ACPs and thus are assigned no-traffic

-guides. There are at-least a dozen such cases within the EPZ; for example, the. intersections'along Mill Road in Hampton'and North.Hampton. In.the absence of traffic guides, it is extremely unlikely.that the optimal signal. assumptions will be observed. In my opinion, these two problems will, beyond any

' reasonable.do'ubt,=cause the ETEs in a real evacuation for some

-regions within the EPZ to be longer than KLD calculated them to be.

We ran I-DYNEV assuming signal timings would be approximately equal'for all approaches, again holding all other inputs constant, and'ETEs increased 45%. However, because I-DYNEV does not allow signal timings to vary throughout the ,

evacuation, we were unable to test directly the more likely effects of non-optimal signal timings on the evacuation times

'of specific subregions under the general evacuation scenario, i

'For example, the intersection of Lafayette Road (Rt.1) and Atlantic Ave. (Rt.1010) affects the evacuation time for the 5 )

mile ring. If the westbound Rt.101D traffic is given more than the 20 seconds (out of a 75 second cycle) allocated in the

un 1 l

Volume'6 I-DYNEV'run for Scenario 1, evacuation out of the 5-mile: ring along Rt.1 northbound will be delayed.- Generally, since; specific. timings'are' assumed in the Volume 6 runs, prudent planners would put mechanisms into' place ensuringsthat they are implemented in the plan. There.are no.such' mechanisms currently. 'Without these, there can be no. reasonable' assurance that ETEs as.short as KLD has projected can ever be achieved.

Q.- You referred in your summary to an'I-DYNEV run that you performed which: included together several'of the effects

- which you have discussed separately above. Would you describe this:run?

A. -I1 wanted to test the effect of several of the factors which describe the conditions of a highly plausible evacuation scenario. These factors' included (1) the more realistic beach parking est-imates from the work of Drs. High and Befort, (2) a more realistic. estimate of I-95 on-ramp capacities, (3) the effects of a' realistic proportion ~of Hampton beach area visitors:headingJto alternative destinations, (4) the.effect at selected locations of returning commuters, and (5) the effect of delayed staffing of TCPs where traffic was routed into opposing lanes. We completed this run for Scenario 4 (summer l weekday with rain) and the result was a 26% increase in ETE for Region 1 (an evacuation of the entire EPZ): instead of taking 9:45, as KLD has estimated, we found the ETE to oc F2:35.

'Again, I want to stress that other than changing I-DYNEV to include these five factors, all other I-DYNEV inputs remained

.the same.

l l J

= _ _ _ _ - _ . . - -. . _ - _ _ _ __. _ _ - _ _ _ _ _ _ _ -__ - - _ - - _ - _ - _ - _ _ _ _ _ _ - _ _ _

s

'i

, .When we 'tried to complete an analogous I-DYNEV run for -

av

, ' Scenario 1 (summer weekend), however, I-DYNEV produced'a run-time error indicating. overflow of'an integer: variable.

-Since we had'been, denied access to the I-DYNEVfsource code, despite our request for it, we could neither check on the source of this model error nor could we correct.it. We were

'able to complete the run only after removing the inputs that-J .specified~ delayed staffing of TCPs. Even without this effect, however,-the ETEs Were dramatically different from those reported in Volume 6. Evacuation times for Region 1 (the entire EPZ) are over 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months , almost double what is given in Volume 6. Similarly, ETEs for~other regions are substantially higher that the Volume-6 values. We simply do not know how much higher the ETEs would be if I-DYNEV Were capable of computing this run with the additional input factor we were forced to leave out -- the-delayed staffing of TCP's.

Q. Did you conduct I-DYNEV runs other than the ones that-you have described so far?

A. Yes.. In total, we ran approximately 75 different simulations. Figure 4 (next page) summarizes the results of the runs described earlier. Attachment 8 to this testimony provides additional details on these runs. We ran several tests to simply validate the I-DYNEV results reported in Volume 6 and numerous others to test ETE sensitivities to other

-inputs. We did not formally audit the full set of runs listed

))

in Volume 6 for internal consistency and for consistency with the TCP_and ACP diagrams (Appendices I and L) and with the I

_ - _ - _ :- . _- i

,1< '

1l1 mCe].

geW - gg - y{Ch z

a, .

s,

)

s r

u .,

I o

I

(

i e

m ., , , , I ,

\ i 1 T

wi]\

x \\ -

. xx \\\\\

m.

xx\ \

N o,

xg \{\

4 g

4 h

4 n

4 g

4 s

y t

c a

p m

" n n t o n n t e

" i w it i o i c f t

u o c m it a f

_m s

d i

a t

e y

a c

R it o

s i

o u

G it l

r n

o a

u d

R u

e p

m a

i T

l i

S a

n g

I l

t mC s

o i kr p

a g

n E

d e

n ib m

o o

l P a c

r p

o R -

d e h c

P C i

n z a h _

C P i c _

l O l a B e a .

i 5 u e N 9- q e v

B I E i t

a -

n r

t e

lA .

S.go -

5u . g2d= _

E-@ g g&a

- x

's igna yi, 1 _

,5g o :

hi fIi l

~1 e !

routing instructions-(Appendices J and K). We did do some spot !

i L checks and found an inconsistency-with:Rt.101 traffic routing.

~

Q. What was.that inconsistency?

A. According to the routing instructions in Appendix J,

'all those.who are routed along Rt.101C are directed onto I 1 north at the I-95/Rt.51/Rt.101C interchange. The TCP diagram

]

for this intersection-(Appendix 5) in fact shows barricades I

" discouraging" westbound movement along 101C beyond the I-95 1

interchange. However, the I-DYNEV simulation for Scenario 1 (summer' weekend)?used in' Volume 6 actually assumes that some; :

traffic continues west along 101C through this intersection, l

despite'the barricades and despite the' routing instructions. 1 Q. ~Did you find any~other discrepancies between what was modeled and what was reported?

A. Yes. We.also.found that the input file provided to us by KLD, which they describe as the input to the Region

'1/ Scenario 3 I-DYNEV run (an evacuation of the entire EPZ on a summer weekday), gives different results than those reported in Volume 6: the inputs that we were given result in a Region 1 evacuation time 40 minutes longer than reported in Volume 6.

The number of inputs required to make each of these runs is substantial, and it is not surprising that there are some discrepancies. The prudent and appropriate procedure for verifying modeling which is the basis of any important planning project is to. commission an independent audit of the work. In this case, independent experts in evacuation planning, traffic e u_______=______z____________________ _ . - _ _

i

. engineering and. computer.modeling1should verify all important

,1 ' inputs and(determine' consistency between'as-modeled (in i I-DYNEV),.as-described 1(in Volume 6), and as-exist (in the L . . actual road-network) conditions. To'our knowledge, no such

' comprehensive review of the work reported in Volume'6 has been i l

completed; 'if it had, at a. minimum the discrepancies reported above would.have been identified. :The fact that no such audit has been performed on a modeling project as important as this one -- where lives-are at stake -- represents a significant deficiency.of the current Plan.

O. Volume 6 describes, in effect, a transportation management plan for the evacuation, and you earlier commented

on some specific deficiencies of that plan. Do you have a professional l opinion as to the " workability" of the overall transportation plan?

A. Yes. The plan in my professional opinion is unworkable. I have strong concerns about the ability of returning commuters, buses and emergency vehicles to pass quick 2y and safely through the Access Control Posts and Traffic Control Posts, as they are currently structured and staffed.

The specific contentions addressed by these concerns include Town of Hampton Contention III/ Basis (C)1 (" gridlock between evacuating private vehicles, commuters and emergency vehicles attempting to enter the EPZ"), SAPL Contention No. 31/ Basis 7

(" commuters will be returning home partly against the direction of evacuating traffic, through intersections with cones blocking desired turning movements") and SAPL Contention No. 37

_ - _ _ _ _ = _ _ _ _ _

("no assurance that effective use of these vehicles will be possible in view of a potential outgoing flow of evacuating

[

traffic"). l

}

My concerns relate to these contentions and are based on several observations and analyses. First, I am concerned

-that substantial delays to vehicles legitimately attempting to enter.the EpZ will occur at Access Control Posts. According to volume 6 of the NHRERP (p. 9-1),

Entry [through the ACPs) should be permitted for the following groups:

Commuters returning to the EPZ, to gather members of their household for the purpose of evacuation.

Transit vehicles (buses, vans, ambulances) dispatched to the EPZ to participate in any evacuation.

All vehicles transporting emergency response personnel.

All other travelers seeking to gain entry to the EPZ should be" denied access and provided with local diversion routes. [ emphasis added]

In order to ensure that "other travelers" are " denied access," New Hampshire Radiological Emergency Response Training 1'

Module 13 (Traffic Management) states that access control managers should allow entry of " emergency response personnel with reasonable identification of the agency they represent . . . members of the press with press credentials . . . area residents with a bona fide i l

need . . . residents must show appropriate identification. ..

[ emphasis added)." Clearly, these people must be stopped at ,

l the access control points to determine that appropriate j l

1 l

screening tests are met. The latter group alone (area residents) will include, according to data in Volume 6, up to 23,000 vehicles driven by EPZ residents who work outside the EPZ and will return to secure their homes and collect' family members. This will result in a need for over 600 resident vehicle identification checks at each post, even if they are spread uniformly across the 38 access control points.

We used data from the Social Data Analysts' survey to determine the likely number of vehicles that would pass through particular critical Access Control Posts. According to these data, the I-95 Post GT-2 (Vol. 6, p. L-5) may need to process up to 5500 returning commuters' vehicles. Assuming that a vehicle identification check requires only 15 seconds on average, almost 23 person-hours of control manager time will be required ht this point just to accomplish this function. Since only 2 traffic guides are stationed on I-95 at this location, this would amount to a delay of over 11 hours1.273148e-4 days 0.00306 hours 1.818783e-5 weeks 4.1855e-6 months in total to these vehicles. Even if identification checks were entirely idaadocci, all of this traffic would still be delayed because

' 1 it is routed via barricades into a single travel lane. At ACP NW-2 (Vol. 6, p. L-29) on the Spaulding Turnpike in Newington, I i

New Hampshire, almost 1900 commuter vehicles (not counting 1

those from Pease Air Force Base) could be returning into the BPZ. With two of the three guides assigned to this location doing nothing but vehicle checks, over 4 and 1/2 hours will be i l

required for this function.

I l

o i

L ,

~"

The effect-of vehicle identification checks and of 1 ' lane restrictions forLvehicles attempting'to enter the EPZ will be to substant'ially delay the evacuation of residents'by

~

delaying the return of commuters. Equally important, there.

will be long. queues,of vehicles awaiting identification checks,

- or simply waiting to-pass:through at'the ACPs.. Such queues will interfere significantly with the movement of' emergency vehicles and buses attempting to reenter the EPZ-and could also directly impede evacuating traffic flows.

Another, serious problem with.the traffic management plan concerns the way in which cones and barricades-will be .

placed at ACPs and TCPs. Returning commuters, emergency buses, and other emergency vehicles traveling in directions other than outbound along the primary evacuation routes will be significantly impeded by these obstacles. At most of these

- control points, each lane through which movement is

" discouraged" is blocked with at'least 3 cones (see, for example, ACP NW-1 on p. L-28). (Attachment 9). The maximum roadway lane widths cited in Volume 6 (p. 3-7) are 12 feet (some lanes are as narrow as 10 feet). If two of the cones are placed at.the outer edges of the lane and the third is pir.ced at the center of the lane, as shown in Volume 6 (see. Attachment 9.) , clear aisles less than 6 feet wide remain. Six feet is barely enough space for a passenger car; school buses and maintenance vehicles are up to 8 feet wide and thus would not fit through these aisles, nor would a passenger car that must turn through the intersection. Instead, these vehicles or barricades are moved.

h---- ---m_a_ - - . - -_...--.m_ -

Furthermore, barricades are specified at several locations where. placement according to the schematics provided in Appendices I and L would physically obstruct traffic flowing into the EPZ. For example, TCP D-HA-02 (p. I-39) (Attachment

5) at the intersection of I-95/Rt.101C/Rt.51 in Hampton shows four barricades across Rt.101C. Given that Rt.51 is blocked in Exeter to westbound traffic by barricades and a " truck,"

Rt.101C serves as the primary route into the EPZ from the west and is, in fact, a designated bus route. Volume 6 states that all barricades will meet MUTCD specifications (p. 9-14). These specifications, listed on p. 9-19 of Volume 6, state that the minimum allowable rail length is 2 feet (4 feet for Type III).

If these four barricades are placed as shown in the TCP D-HA-02 l diagram.(Attachment 5), there would clearly be insufficient room for either cars or buses to proceed westbound along Rt.101C as would be required for orderly movement of these incoming vehicles.

A more general overall concern that I have with the descriptions of traffic control strategies at the ACPs and TCPs is that there is substantial ambiguity and, in some cases, seeming contradiction as to how the controls are expected to work. In some cases, such as the one cited above at the I-95/Rt.101C/Rt.51 intersections, traffic is apparently to be allowed past barricades. This is consistent with the policy stated in Volume 6 (p. 7-1):

1 We employ the terms " facilitate" and

" discourage" rather than " enforce" and

" prohibit" to indicate the need for flexibility in performing the traffic control

function. There are always legitimate reasons for a driver to prefer a direction other than that indicated.

'However, there are several locations where allowing a driver to move in a " direction other than that indicated" will i result in a head-on conflict with evacuating traffic. For example, a " movement discouraged" indication is given on TCP l V-AM-06 for the I-95 southbound off-ramp traffic (p. I-19)

(Attachment 3). However, as noted above, this of f-ramp ha s been converted for use as a primary evacuation on-ramp. If the ~ j off-ramp traffic is not prohibited, a direct head-on conflict will be created. Similarly, at TCP F-EX-04 (Rt.101C/Rt.51/Rt.101D in Exeter), movements from Rt.101C and from Rt.101D onto Rt.51 eastbound are " discouraged" ) (p. I-74)

(Attachment 10). However, unless these movements are prohibited, a head-on conflict with evacuating traffic will be created on Rt.51 at the I-95 interchange (p. I J9) (Attachment 5). These head-on conflicts will, at a minimum, be disruptive to traffic flow and will, additionally, create substantial safety risks.

Q. What will be the effects on ETEs of all of these traffic management problems?

A. It is difficult to say precisely what the effects will be, but I can say qualitatively that the effects will include (1) substantial delays to emergency vehicles, buses and ;

commuters who are legitimately attempting to enter the EPZ, (2) substantially increased risks of vehicular accidents, l I particularly at key bottleneck locations, and (3) reduced l 1

l

. efficiency of evacuation flow. None of these effects have been modeled directly in the analyses presented in Volume 6.

Q. Have you had an opportunity to review the analyses presented in Volume 6 related to provisions for transit-dependent individuals and the corresponding contentions?

A. Yes. SAPL Contention 31/ Basis 6 asserts that the telephone survey data used by KLD to develop estimates of transit-dependent populations is biased by non-response and therefore the data are not reliable. Basis 12 criticizes other

, data used by KLD to determine bus mobilization times, Basis 13 specifically rebuts the KLD estimates of transit-dependent population; and Basis 17 states that KLD's adjustments for out-of-service vehicles were applied incorrectly. SAPL Contention 37 asserts that buses (and other emergency vehicles) will be significantly impeded in their trips into the EPZ. p KLD's analyses begin by using the First Market i 1

Research telephone survey data to determine the number of people in each town who either do not own a car or who do not have access to a car at the time the evacuation is ordered.

These numbers are increased by 6% to account for vehicles which may be out-of-service and are further increased to account for statistical sampling error. The resulting estimates of people requiring transit are then reduced by half to account for the assumption that 50% of the earless evacuees would find someone with whom to carpool, Bus loading assumptions are used to compute the required number of buses and, finally, ETEs for these buses are computed based on KLD's estimates of bus l

1 L _ _ _____--- _ _

o L mobilization'timesfand route travel times.

Q. .Do you believe that the Volume 6' estimates'of transit.

dependentLpopulations are reasonably reliable?

A. No. ;As stated clearly in Dr.: Cole's testimony, the' Fi'rstJMa'rket Research survey,.from whi~ch KLD drew primary data, is generally: unreliable. Further, the KLD analysis,makes.

1several unsupported assumptions about the conditions under whi~ch cars would be available for use in an evacuation. In i

addition to these general problems, the adjustment for out-of-service vehicles was, applied incorrectly, as asserted in SApL)31/ Basis 17, and the assumed extent'of.carpooling is not-substantiated.

Briefly, the : adjustment f actor for out-of-service

.veh'icles was applied in Volume 6 by increasing the transit-dependent population by 6%. The stated assumption (p, 11-8)-is that "the average vehicle is out of service three m weeks out of the year." With this assumption, KLD derives its 6% adjustment factor as follows: 3 divided by 57 = 0.058. In i

. fact, if 6% of the vehicles were out-of-service, as assumed, ;

theLeffect on the size of the transit-dependent population would be'much larger.- A simple example illustrates this

. point. Assume that a town has a population off15,000, that an !

average household size is 3 people (5,000 households), that 1 1,000 households have no cars, and that the rest (4,000 households) have 1 car (and that that car will be at home at the time of the evacuation). With no vehicles out of service, 3,000 people will be transit dependent (= 1,000 hhs x 3 .

per/hh). If 5% of the cars owned in the town were out-of-service, an additional 200 households (= .05 x 4,000 p hhs) or 600 individuals will be transit dependent, a 20%

increase over the normal transit-dependent population. The method used in Volume 6 would increase the transit-dependent population by only 5%, incorrect.y assuming that the out-of-service fraction can be applied directly to the transit-dependent population.

~

In his March 25 affidavit (SAPL 31), Mr. Lieberman defends the " deliberately selected conservative figure of 6 percent with which to estimate persons needing transit because, of out of service vehicles" as exceeding what "available sources indicate, which is . . . 1.1 percent of the time." (p.

25) He does not, however, defend the subsequent misapplication of.this percentage. Lieberman offers, as candidate populations for the adjustment, either the "EPZ population who own one vehicle" or the " estimated transit-dependent population" in his

{

evident uncertainty as to which is the more appropriate. Id.

In fact, neither is correct. Proper, conservative adjustment for the increase in the transit-dependent population arising from out-of-service vehicles mandates the consideration of the following populations: (1) those in households with one car available, where all household members generally have access to the car, and (2) those who normally drive in the cars in one-car housaholds where not all members generally have access.

The assumption that 50% of those not having access to a car will carpool is critical in the Volume 6 estimate of

l transit-dependent populations. Volume 6 notes (at p. 11-8) that nearly'80% of the transit-dependent population ride-shared in the Mississauga, Ontario, evacuation. However, the l

comparability of that observation in terms of the general conditions of the evacuation, geographic distribution of the population, vehicle occupancy and annaunced availability or unavailability of bus transportation is not formally evaluated. The Seabrook EpZ is characterized by a large number of transients in the evacuation stream; the effects of these transients on the prop nsity to ride-share should be evaluated more carefully. In general, since the number of buses required is directly dependent on this calculation, a more reliable basis should be provided.

Volume 6 " verifies" the transit-dependent estimates by comparing them to the results of a June 1986 NHCDA survey. That survey however is, at best, at equally unreliable instrument; the non-response bias from the mailout/mailback survey format renders it unsuitable for this enumeration task.

The Social Data Analysts survey described by Dr. Cole ;

illustrates a more direct method of determining the number of people who would require public transportation in an evacuation. A simple question was included in the !

questionnaire: "How many individuals in your family will need public transportation in order to evacuate?" We tabulated the responses to this survey question and expanded these to the full EpZ. The results are, generally, a transit-dependent population twice as large as that estimated in Volume 6 (see

1 Figure 5, next page). These data indicate, therefore, a need for-at least twice as many buses as contemplated in Volume 6.

l The effect.of having only half the number of buses needed would' f

be to substantially increase the ETEs for transit-dependent I individuals above those presented in Volume 6, Table 10-10, i 1

pp. 10-71 through 10-74.

Q.

Are there other factors that would also increase ETEs for transit-dependent individuals?

A.

Yes. In SAPL 31/ Basis 12, KLD's estimates of bus mobilization timet are challenged as being overly optimistic.

According to pp. 11-18 of volume 6, these estimates of bus mobilization times were derived from a " telephone survey of organizations which own and operate buses." Volume 6 states (pp. 11-19) that based on this telephone survey, 50% of the available buses could be mobilized in one hour. However, a 1987 telephone survey conducted by NHCDA reveals that only 30%

of the required vehicles could report to staging areas within one hour (March 25, 1987 Affidavit of Edward B. Lieberman (SAPL 31), p. 21).

In fact, the " bus mobilization time" study relied upon in Volume 6 was apparently carried out in a haphazard manner. Concerning the bus mobilization time survey, the Response to Massachusetts Attorney General Interrogatory 52 states:

Because of the limited amount of information which was sought, there was no need to develop a formal survey instrument or methodology.

The importance of having accurate bus mobilization 1

times to an accurate estimation of ETEs for transit-dependent L-_-______ . _ _

Figure 5

)

[ l i

1 l 2500 Persons Requiring Transit NH Plan Estimates New Estimates 20@ -

/ -

j h 1

$ 1500 -

z 3 i 3

3 1000 -

/

f h

.i..l 5 5 5 ih 5 $$

k l .

5 ld.a. hhd..h

$ k 5 h h hco '

M $

$ u1

.5 g d o a .g .o 'S E

{# o k$yl* AN1 03 '

1 8 2 z Sources: Resource Systems Group. SocialData Analysts vs. ,

NHRERP Volume 6

- 64a -

n:

peopleLwould suggest:the need for more formality and reliability?in the data collection-process.. The Response,goes on.to state that: a To th'e best of KLD's recollection, this survey was conducted in the spring of 1986.

Despite difficulty in recalling when this vital information was obtained, the Response describes how KLD' conducted this survey:

While there was no way to check the.

reliability of these returns, KLD did-engage in extended discussions over this matter with each respondent.- During the' course of -

-these discussions, KLD emphasized the need for accurate and reliable information and always fed back the~ responses obtained to confirm that the_ respondent felt comfortable with the information that heoor she gave.

KLD also ascertained that the person was either'a.,'anager or a responsible employee of the firm to ensure that the information was reliable.

This rather: formal sounding engagement in " extended I discussions" on reliability should have been coupled with a formal survey.

An equally important problem,' discussed in more detail above, is the likely substantial delays and queueing at several of the important ACPs due to vehicle identification checks and lane restrict'i'ons. The further difficulty that buses will have t

in maneuvering through cones and barricades at each of the many TCPs encountered in each of the routes will further impede inbound bus flows. However, Volume 6 assumes an average speed for incoming buses of 40 to 50 mph. No specific basis for this '

calculation is.given, though Volume.6 states that the average inbound travel distance is 30 miles (p. 11-20). Assuming that

{

l 1

/,V i

\

x <' g l T - j only one-fourth of thst distance .(7.5 mi.) is within the EPZ

~

c' 1

and the remainder (22.5.mi.) is traveled at 55 mph, the posted i j

l

~ speed limit on most major access roads,'in order to average 50 f !

mph, the EPZ travel would have to be'at speeds averaging 40 g mph. '

The likelihood of achieving that average speed inside the EPZ, and thus of providing transit-based evacuation in'the times assumed inqvolume G, is remote given the many, impeding 4

q ,

factors that wia74 ,. be present under evacuation condt,ttons. .

y.

Q. You have(teported in this testimony 4many examples of i; a

changes in Il0NY$V inputs which dramatically affect the ETEs; a

yet the " sensitivity tests" described in Volume 6 and those reported in Mr. Lieberman's March 25 affidavits show generally

.only small effectsjin ETEs. Scw do you explain these differences?

A. Very simply, we evaluated different sets of

. conditions. M .

Lieberman modeled those which he apparently thought were.most likely and we modeled conditions that we have demonstrated through this and other supporting testimony are more likely. iCpmputer models of behavior as complex as an i

evacuation from the Seabrook EPZ understandably involve many assumptions about values of input parameterc. However, where these values are uncertain, it is important to determine whether changes in the assumed values will affect the major conclusions'that are drawn from the work. In the case of the

]

ETE study reported in Volume 6, there are many assumed model inputs'with which there is reasonable uncertainty regarding the l

"true" value. '

l 1

L_-____--- l

h<

q J

l volume 6 reports sensitivity tests for only four sets of conditions (pp'. 10-15 to 10-19). The first set deals with differentLassumptions about beach population; however, the two additional scenarios described look only at the effect of reductions in beach population to 80% and 60% of the original estimates.. The work reported by Drs. High and Befort indicated b that much. higher beach populations are both possible and likely

~

to occur. Two additional sets of KLD's sensitivity tests deal with diffeient rates of accident escalation. The final reported set of sensitivity runs describes the effects of two sets of highway, improvements. Noticeably lacking are tests which describe the effects of reduced capacity. In general, the KLD sensitivity tests ignore important issues such as the

'effect of returning commuters, the effect of evacuees choosing to travel.to host locations other than those assigned, the effect of a possible underestimation of beach population, and the effect of spontaneous evacuation outside the EPZ. The

/ .

March,2S, 1987,,Lieberman affidavit cites tests of the effects l

of "10 road blockages at different locations throughout the l

~

EPZ."- These " blockages," however, were specified to be of limited duration and served to reduce capacity over that period but not to " block" flow. {

I There are also several effects that KLD chose not to 1 model explicitly,: arguing instead that they would have little impact on ETEs or that they are outside the scope of the work.

~

Some of the effects KLD ignored are: (1) the substantial flow ,

l of returning commuters, (1) the delay caused by identification l I

l I

l

- _ - _ _ _ - - i

)

checks, (3) the impedence produced by cones and barricades, (4) i i

the choice by beach evacuees of alternative destinations, (5) I the "non-optimal" traffic signal timings, and (6) the area's

]

projected growth. Each of these effects has an influence on ll

)

ETEs and should be evaluated through sensitivity tests such as the ones reported here. 1 Q. Have you reviewed Appendix 4 to NUREG-0654/ FEMA-REP-1, Rev. 1 ("NUREG-0654"), which pettains to the requirements of an evacuation time assessment study?

A. Yes, I have. I understand that it presents "an example of what shall be included in an evacuation times assessment study and how it might be presented." (NUREG-0654, App. 4, 4-1).

Q. Does Volume 6 of the NHRERP, Rev. 2, meet all of the requirements set forth in Appendix 4?

A. No, it does not. For example, in paragraph I.A. (at

p. 4-1), Appendix 4 call for a " vicinity map" which identifies, among other things, " topographical features." However, there are no maps in Volume 6 which identify topographical features.

In paragraph I.B. (at p. 4-2), Appendix 4 states:

"All assumptions used in this analysis shall be provided." As my previous testimony has recounted, during our assessment of KLD's work we uncovered assumption after assumption which was not provided in Volume 6. One example is the assumption that KLD makes, but does not mention, about returning vehicles -- that they will not impede the efficient flow of evacuating vehicles out of the EPZ at critical intersections.

F .i ?r n .,

, s-- Q- f f i +

1 ,

s.' Another example is that KLD does not' explain its assumptions or

'N .

.e , s method of interp>1ating its liata to determi'ne tl'e "end time,"

i.e...the ETE, for each of its model tuns. There are numetous g important assumptions made by KLD which are not reported in L . b.

Volarc.e L .

Paragraph b.C. ( " Met h o do.'.o gy " ) in Appendf.x 4~ (at 4-2)

, +

states as follows:

A descr'ipt' ion of the method cf analyzing the evacuation times shall be providsd. If ^

computer models are used,, a general description of the algorithm shall be provided along with a source for obtaining further information or documer..tation.

l As I have noted, the descriptl>n contained in Volume 6

,m 'of the method used to calculate evacuation times was incomplete. While a tary general description of- I."DYNEV was contained ir. Vrdurac 6, it did not offer enough information for ud (or anyone else) to condact a meaningful'ussessment. .Thus, we turned to the " source," i.e., KLD Associat'as, to obtain

.further information and documentation. While we did obtain a copy of ithe I-tYNEV system itself, when we requested a cooy of tne " source. code," i.e., the primary computer program upon

/which I-DYNEV is,' built, KLD refused to provide us with this

, documentation. In my professional experience, whenever modeling projectre have been undertaken which are meant to come -

s

, undet government scrutiny, this primary computer program is generally :nede available to those reviewing the model and its L. Applicationonagive.cprcject. I assumed that in r'equiring 3 a

l' " source for obtaining f41ther information or documentation,"

NUREG-0654 was implicity requiring that this documentation then

) .;

r v l'

~

69 -

x be made'available by the " source." Here, perhaps the single.

most' critical piece of doc'umentation, I-DYNEV's source code, was requested, rand the request was refused. This is not the L

only documentation we requested:which was not available. As I

,have noted, KLD discardedfall the_ original. documentation from 7

the' extensive highway system' field survey lit says it

. conducted.' See Vol. 6, p. 1-10.

It also discarded the

. sketches it made of each of the August 11, 1985,' aerial

. photographs it examined in. counting the beach area parking

. capacity.

It is'not apparent from the slides themselves what

'KLD counted as a parking spacer only the sketches KLD made of these slides would document where it claims these spaces do and do not exist.

Yet these sketches were unavailable to us as well.- NUREG-0654's requirement is meaningless if all it means

'is that the " source for obtaining further information" must be identified but that no documentation need be provided by the

" source."

Further paragraphs of Appendix 4 to NUREG-0654 pertain to " Demand Estimation" (par. II), " Traffic Capacity" (para.

III), and " Analysis of Evacuation Times" (para. IV). As I have stated in my testimony, I believe that KLD has significantly underestimated demand, overestimated capacity, and, as a result,' underestimated ETEs. For this reason, even if the Volume 6 ETE study met the checklist of requirements set forth in Appendix 4 to NUREG-0654, which I do not believe it does, it would not ensure that the ETEs are reliable or accurate to a reasonable degree. But if that checklist of requirements had

_ _ - - _ - - - - - - - ~

I 1

'been? met,;then our review of the NHRERP's ETE j

~

stddy would have

.been much-easier to conduct.

'l Let'me give you an' example.

Appendix 4 to'NUREG-0654'(at~p.

In section.III (B.) of y 4-5), which pertains'to " Road Segment. Characteristics," there is a'requiremen t ctable~such'as example Tabl that "[a]

e 1 shall be provided indicating all-the evacuation route segments and their charact eristics, i

ncluding capacity."

The referenced Table 1 (on p. 4-15.of NUREG-0654)

(Attachment 11,. hereto) indicating'" comments." contains a' column for- "

A footnote.gives this instruction-

'regarding.what ,

" comments" to included:

" Indicate any special ,

conditions that may affect roadway capacity."

In doing our review of KLD's work,-we would haveound f it-extremely helpful to have had this information f or each

-roadway' link.

But nowhere in Volume 6 or anywhere else:is this information provided.

This information was among the data

. collected by KLD staff when, in August 1985, it drove the entire highway system within the'EPZ and recorded the

. characteristics of each section of highway, including special conditions that.may affect capacity.

See Volume 6, p. 1-10.

'Because this.information was not in Volume 6 KLD for

, we asked this1information and were informed that the apes audio on t which.this information was recorded had all n been re-us

-that no transcript or other documentation of this roadway data existed.

KLD said that they had adjusted I-DYNEV's link

. capacities accordingly, based on this data ,

but having so programmed the model, it had no further need for the rawa.dat

Thus, instead of providing the detailed road segment

. characteristics NUREG-0654 requires, KLD provided only a table, in Appendix N to Volume 6, which lists KLD's estimates of-capacity for each link. For us, this meant that when we wanted to check KLD's capacity estimates for any given link we had to go out to the EPZ and drive the link ourselves, gathering the very same road characteristic data that NUREG 0654 requires be included in the Volume 6 ETE study. This made it extremely j time consuming to reveiw all of KLD's capacity estimates. As a result, we drove only those links we found to be most critical, checking for "special conditions" which might reduce capacities. As I have noted, we found some important "special conditions" which apparently were not considered in KLD'S analysis.

There is another sensible reason for NUREG-0654 to have this requirement that special roadway conditions ~affecting capacity be listed for each link. As conditions change on any link, e.g., as road improvements are made, or as impediments such as construction projects occur, there is no simple way of knowing whether these will affect the ETEs in the absence of "special condition" data. For example, if a narrow bridge in the EPZ is widened, those responsible for keeping the ETEs up-to-date will not know whether that may affect (lower) the ETEs or not (for that town or region), since they do not know whether KLD considered the narrow bridge to be a factor l

limiting capacity at all when it did the ETE study. If those l

responsible for the plan had a simple list of each link's i r ,

4-

"special conditions"1affecting' capacity as they monitor and

. record roadway characteristic changes in the EPZ, they could reference ~that list'and.know whether to update the ETE for that.

town (or ERPA). As it stands now, they will have no idea, as road-characteristics change, whether.to update'that ETE.

Prudent planners would always have available the "special condition" data that is absent from this Plan. The simole 1 1

solution to this. serious defect in the ETE study is to~ require l

that the "special condition"' roadway characteristic data be reported in the way required by NUREG-0654, or any similar way, i even.if this.means re-driving each link in the EPZ evacuation network-to regather the data.- I

.I Q. Do you feel that you.have been able to complete a thorough evaluation of the reliability of the data and analyses on which Volume 6 is based? f 1 A. No. We have been able to evaluate many, but not all, of the data and analyses reported in Volume 6. An initial i

necessary_ test o f the validity of any computer-model-based study is that the results can be independently derived and f J

verified-(see, for example, Meadows and Robinson, The Electronic Oracle: Computer Models and Social Decisions, John .

I Wiley & Sons, 1985). Documentation should be sufficient to allow an independent analyst to audit the trail leading from the data and assumptions to the actual inputs used in the study. Moreover, the computer model itself should be available for full scrutiny and should be documented to the point that the basis for each key model equation is described and its

computer language implementation (source code) is revealed.

The model outputs should also be available for comparison to independently derived outputs. The information provided in Volume 6 falls far short of this necessary level of documentation. This is understandable since the purpose of Volume 6 is simply to report results of the study and to point the reader to supplemental sources of data. However, even after obtaining responses to interrogatories directed specifically at obtaining the information necessary for our .

work, we_found sig7ificant gaps in the audit trail, for example:

Records of KLD field data collection efforts documenting the critical physical attributes of the area's roadways, including both audio cassettes and typed transcripts referenced in Volume 6, have been destroyed by KLD staff; No written records were maintained of a survey conducted by KLD to determine the availability of buses for evacuation of transit-dependent individuals; We were denied access to the computer source code on which I-DYNEV is based. This meant that we were unable to verify that the model correctly implements the methodology described in reference documents. It also meant that we were unable to determine the i source of'" suspicious" model behavior such I as a seeming over-representation of ramp {

capacity under congested conditions and a 4 run-time error which prevented us from {

running one of our key model simulations; i de were given conflicting data by KLD; the '

computer input file that supposedly was used for the Scenario 3 run reported in volume 6 does not produce the output described 4 therein. {

These gaps, among several otherc, impeded our efforts, and would impede anyone's efforts, to fully critique the work.

l

More importantly, they substantially diminish the ability of anyone other than the same KLD staff who produced the current version of Volume 6 to update the. work in later years, as required by NUREG 0654. For example, as noted above, if future road improvements are made to remove existing bottlenecks, it will not be clear whether those bottlenecks were considered in KLD's rating of a particular roadway. Or, if a bus company goes out of business, it will not be known whether that company was relied upon in KLD's survey. Further, others without access to the I-DYNEV model's source code will likely be unable to extend the runs reported in Volume 6 and will not be able to determine the source of model anomalies.

Q. Do you believe that Revision 2 of Volume 6 accomplishes its stated objectives?

A. No. As stated on p. 1-1 of Volume 6:

This plan is designed to protect the health and safety of the public in the event that an emergency evacuation is ordered as a protective action in response to an accident St Seabrook Station.

The material contained in Volume 6 undoubtedly represents a significant amount of effort expended by KLD and its subcontractors to address this objective. However, the insistence throughout that document that the ETEs produced reflect " conservative" assumptions is challenged by the data and analyses presented in this testimony. Generally, the assertion that ETEs can be estimated with a reasonable degree of certainty from the data and analyses presented in Volume 6 is not supported. Further, the transportation management plan

has-significant deficiencies which will reduce the efficiency of the evacuation and which could, in practice, expose evacuees to risks of-nerious traffic accidents.

l There are three major sources of error or uncertainty in estimation of the evacuation times reported in Volume 6: in calculations make by I-DYNEV, in the quantitative values of inputs to I-DYNEV, and in the overall assumptions made about the behavior of evacuees and of evacuation personnel. This testimony has dealt primarily with the second of these three sources, the values of inputs to I-DYNEV. In some cases, such as for returning commuters, we have described effects which ,

could be more precisely quantified given additional data collection and/or analysis. 17 the case of the beach population, we have defined, by re-estimating the number of parking spaces, a wider range of vehicle populations that could be present on any given summer day. Since no one would know precisely how many beach visitors actually occupy the beach i area on the day of an emergency, there would correspondingly be a wide range of uncertainty in evacuation times for this population. This uncertainty could be eliminated only by conducting daily aerial overflights of the area, but it could be reduced substantially by compilation of a statistical data base that would correlate actual beach counts with weather conditions, daily / weekly trends and yearly growth factors.

Of the other two major sources of uncertainty in the evacuation time estimates, it is my professional opinion that behavioral assumptions are more significant than is uncertainty 76 -

resulting- from the methods used in Z-DYNEV .

Having conducted extensive research on travel behavior aving reviewed andtheh testimony of Drs. Zeigler, Johnson Cole, and Luloff I am concerned-about extrapolating, as the pl very simple models of human behavior to sitan implicitly does well outside the realm of common experi uations that are ence. My fundamental concern is that for those in the beach a evacuation times presented both in Volum reas of the EPZ the testimony are so long that e 6 and in this an orderly evacuation may no longer applythe P Certainly, the evidence provided by KLD does not substanti ate these dubious assumptions and thus does not provide reas an orderly evacuation, under these circum tonable assurance tha Q. s ances, would occur.

Is there any other testimony that y present at this time? ou would like to A. -No. Thank you.

i I

l s

\

UNITED STATES OF AMERICA fUCLEAR REGULATORY COMMISSION Before Administrative Judges: 4 Helen F. Hoyt, Chairperson Gustave A. Linenberger, Jr.

Dr. Jerry Harbour

)

In the Matter of )

)

PUBLIC SERVICE COMPANY OF NEW ) Docket Nos.

HAMPSHIRE, ET AL. ) 50-443-444-OL (Seabrook Station, Units 1 and 2) ) (Off-site EP)

) September 14, 1987

)

ATTACHMENTS TO TESTIMONY OF THOMAS J. ADLER ON BEHALF OF THE ,

ATTORNEY GENERAL FOR THE COMMONWEALTH OF MASSACHUSETTS ON REVISED SAPL CONTENTION NO. 31, SAPL CONTENTION NO. 37, AND TOWN OF HAMPTON REVISED CONTENTION III TO REVISION 2 (THE "ETE CONTENTIONS")

Department of the Attorney General Commonwealth of Massachusetts One Ashburton Place Boston, MA 02108-1698 (617) 727-2265

UNITE'D STATES OF AMERICA NUCLEAR REGULATORY. COMMISSION Before' Administrative Judges:

Helen F. Hoyt, Chairperson '

Gustave A. Linenberger, Jr.

Dr. Jerry Harbour

)

In the Matter of )

)

PUBLIC SERVICE COMPANY OF NEW ) Docket Nos.

HAMPSHIRE, ET AL. ) 50-443-444-OL (Seabrook Station, Units 1 and 2) ) (Off-site EP)

) September 14, 1987

)

ATTACHMENTS TO TESTIMONY OF THOMAS J. ADLER ON BEHALF OF THE ATTORNEY GENERAL FOR THE COMMONWEALTH OF MASSACHUSETTS ON REVISED SAPL CONTENTION NO. 31, SAPL CONTENTION NO. 37, AND TOWN OF HAMPTON REVISED CONTENTION III TO REVISION 2 (THE "ETE CONTENTIONS")

. I Department of the Attorney General .

Commonwealth of Massachusetts l One Ashburton Place -

02108-1698 Boston, MA ,

(617) 727-2265 1

1

ATTACHMENTS Attachment 1 Professional Qualifications of Thomas J. Adler Professional Qualifications of Colin J. High Professional Qualifications of Dennis L. Meadows Attachment 2 SAPL Contention 31 SAPL Contention 34 SAPL Contention 37 Town of Hampton Contention III Attachment 3 Traffic Control Post for Rt. 110/I-95 Attachment 4 Traffic Control Post for Rt. Il0/I-95, with modifications Attachment 5 Traffic Control Post for I-95/Rt. 51/

Rt. 101C Alcachment 6 Traffic Control ?ost for Exeter on Rt. 51

' Attachment 7 Traffic Control Post for Rts. 1, lA and 110 Attachment 8 Description of Sensitivity Runs

' Attachment 9 Access Control Post in Newington

' Attachment 10 Table 1 (Roadway Characteristics) from Appendix 4 of NUREG-0654 Attachment 11 Description of ETE Scenarios Used in Volume 6 Attachmenc 12 Description of ETE Regions Used in Volume 6

,---.----------,-------,- ------,------- -. r--

u.

k ATTACHMENT 1 1

l l

l l'

R6 sum 6 THOMAS JAY ADLER Resource Systems Group Route 5 South P.O. Box 1104 Norwich,VT 05055 802-649-1999 Born: August 22,1950; Portland, ME EDUCATION B.S. Comell University (Civil & Environmental Engineering),1972 S.M. Massachusetts Institute of Technology (Transponation Systems),1975 Ph.D. Massachusetts Institute of Technology (Transponation Systems),1976 PROFESSIONAL EXPERIENCE 1986 - Present President, Resource Systems Group; Principal in Charge for transportation projects.

1984 - 1986 Research Associate Professor of Engineering, Thayer School of Engineering, Danmouth College. Coordinator of Master of Science Program in Resource Systems at the Resource Policy Center. Graduate and undergraduate instruction in transponation, resource policy, systems modeling, and engineering sciences.

Principal investigator on numerous sponsored projects.

1982 - 1984 Acting Director, Resoture Policy Center, and Associate Professor of Engineering, Thayer School of Engineering, Dartmouth College; Adjunct Associate Professor of Policy Studies, Dartmouth College.

1976- 1982 Assistant Professor of Engineering, Thayer School of Engineering, Dartmouth College.

1975 - 1986 Independent transponation consultant for state, regional and local agencies. !

Transportation consultant for Cambridge Systematics, Inc., Kendall Square (

Bldg., Cambridge, MA; consulting on various national transportation research projects. l l

1975 - 1976 Research Fellow, Joint Center for Urban Studies of Harvard and M.I.T.

Involved in interdisciplinary program of urban studies.

1973 - 1976 Research and teaching, Transportation Systems Division, M.I.T. Primary research on U.S. DOT contract to M.I.T. entitled " Experiments to Clarify Priorities in Urban Travel Demand Forecasting Research and Development."

1972 - 1973 Research Assistant, Transponation Systems Division, M.I.T. Research on U.S.

DOT (Transponation Systems Center) contract to M.I.T. to study the computer i requirements of a transportation research agency. i 1971 - 1972 Civil engineering consultant, Carrol E. Taylor and Associates, Auburn, ME.

Responsible for the development of computer software for the analysis of stmetural and waste water collection systems.

1

Adler R6 sum 6 PROFESSIONAL ACTIVITIES Transportation Research Board. Member and Incorporator, Intemational Network of Resource Information Centers (INRIC). Member, American Society of Civil Engineers, American Society for Engineering Education.

AWARDS /:ND HONORARY SOCIETIES Chi Epsilon, Civil Engineering Honorary Society Sigma Xi, Scientific Honorary Society Catherine B. Wurster Fellow, Joint Center for Urban Studies of Harvard and M.I.T.

PUBLICATIONS

" Formulation of Travel Demand Modeling Requirements", with J. Bottom. Working Paper DOT-WP-1, Transportation Systems Division, M.I.T.,1973.

"A Joint Choice Model of Social-Recreational Travel", with M.E. Ben-Akiva. Working Paper DOT-WP-10, Transportation Systems Division, M.I.T.,1975, i

" Comparison of Two Travel Demand Modeling Systems: The M.I.T. Disaggregate Behavioral Models and the Urban Transportation Model System", with M.E. Ben-Akiva.

Working Paper DOT-WP-13, Transportation Systems Division, M.I.T.,1975.

Experiments to Clanfy Priorities in Urban Travel Forecasting Research and Development:

Summary Report, with M.E. Ben Akiva and J. Jacobson. Prepared for U.S. DOT contract with M.I.T., DOT-OS-30120,1976.

Modeling Non Work Travel Patterns. Center for Transportation Studies (CTS) Report, M.I.T., Cambridge, MA,1976.

" Joint-Choice Model for Frequency, Destination, and Travel Mode for Shopping Trips", '

with M.E. Ben-Akiva. Transportation Research Record #369, Transportation Research (

Board, Washington, D.C.,1976. i "Disaggregate Models of Trip Distribution", with S.R. I.crman, in Behavioral Travel-Demand Models. Lexington Books, D.C. Heath and Co., Lexington, MA,1976.

"A Theoretical and Empirical Model of Trip Chaining Behavior", with M.E. Ben-Akiva, Transportation Research, Series B, Methodology,13B, pp. 243 - 257,1979. j l

The Direct andIndirect Costs of Transporting Wood Chips to Supply a Wood-Fired Power Plant, with M. Blakey and T. Meyer; report submitted to the U.S. Department of Energy Solar / Biomass Division under Contract EG-77-C-0204487, DSD #103, February 1978.

" Directions for Improvements in Urban Travel Forecasting Procedures", with M. E. Ben-Akiva and S. Lerman; in Conference Summary and White Papers, U.S. Department of Transportation,1979. ,

i Techniquesfor Analyzing the Performance ofRural Transit Systems, Vols. I & II, with S.R. Stearns and Y.J. Stephanedes, prepared for U.S. Department of Transportation, Research and Special Project Administration under Contract DOT /OS/80006, September l l

1979. )

}

Adler R6 sum 6 "ENTRANS: A Model of the Interactions Between Energy Supply and Transportation Enctgy Use," in Proceedings of the Summer Computer Simulation Conference,1979.

"Disaggregate Models for Decisions Other Than Travel Mode Choice," in New Horizons in

' Travel Behavior Research, W. Brog and P. Stopher, eds.,1980.

" Forecasting Experiments for Rural Transit Policymakers," with Y. Stephanedes, Transportation Research Record #718, Transponation Research Board, Washington, D.C.,

1979.

Interactions Between Energy Supply and Transportation-Related Energy Use, Vol.1, with J. Ison and J. Geinzer, final report submitted to U.S. DOT under Contract DOT /RC/82003, DSD #182, January 1980.

Interactions Between Energy Supply and Transpor;ation-Related Energy Use, Vol. Il Technical Appendices, with J. Ison and J. Geinzer, final report submitted to U.S. DOT under Contract DOT /RC/82003, DSD #183, January 1980.

" Analysis of Long-Term Transportation Energy Use," with J.W. Ison, Transportation Research Record #801,1981.

Analysis of1* transportation Energy Conservation Policies Using the ENTRANS Model, with J.W. Ison, Solar Energy Reseamh Institute report, January 1981.

Guide to Forecasting TravelDemand with Direct Utility Assessment, with G.A. Kocur and others, U S. DOT, Urban Mass Transit Administration, Repon UMTA-NH-11-0001-82-1, September 1982.

Transportation of U.S. Coal Through Eastern Ports, with F.W. Lipfen and others, Report prepared for U.S. DOE under contract DE-AC01-81FE-20036, September 1982.

Technical Review of the ORNL Highway Gasoline Demand Model, report DE-ACO1-81El-11976, U.S. Depanment of Energy, Energy Information Administration, February 1982.

"Long-Term Effects of Auto Fuel Efficiency Regulatbn," presented at TIMS/ORSA Joint National Meeting, Detroit, Spring 1983.

"Use of Microcomputers for Resoume Planning," presented at Second Annual Meeting of the Balaton Group, Csopak, Hungary, September 1983.

Paratransit Planning System PARAS Cost ModelDescription, with H. Hazard, DSD #495, U.S. DOT Report UMTA-NH-06-0002 1, January 1983.

Low Density Transit Planning Package, with S. Tahmosh and M. Bunon, turnkey PASCAL software system, distributed by U.S. DOT UMTA, February 1984.

" Review of Recent Advances in Travel Demand Analysis," Transportation Science, V 18, N. 3, August 1984.

An Analysis of Wood Transport Systems: Costs and ExternalImpacts, USDA Forest Service Report 23-729, March 1985, i A Review of FORECASTS, with D.L. Meadows, prepared for Defense Communications Agency, September 1985.

l L_________. . _ _ _ _ _ . _

l L

Adler R6 sum 6 . COURSES TAUGHT DARTMOUTH COLLEGE:

Graduate Level Transportation Systems Analysis Trans;mrtation Engineering Travel Demand Forecasting Models in the Policy Process Probability and Statistics

- Optimization Modeling Consumer Choices Resource Models UndergraduateLevel Introduction to Computer Science Engineering Economics and Optimization Optimization Applied to Environmental Engineering i Modeling Complex Systems MASSACHUSETTS INSTrrUTE OF TECHNOLOGY:

Travel Demand Forecasting (Summer course for practicing professionals)

SUMMARY

OF MAJOR FUNDED PROJECTS Travel Demand Model Estimation Using Direct Utility Assessment Funded by Pennsylvania High Speed Rail Commission, Peat Marwick and Mitchell, Pinellas and Hillsboro County (Florida) and Florida DOT 1986- present 1

Resource Systems Group is actively working with several clients to develop travel demand estimation models for major transportation projects including high speed passenger rail systems, urban transit systems and ride-sharing programs.

Review of Travel Demand Estimation Procedures Funded by Ohio Department of Transportation 1986 - present Resource Systems Group has an ongoing contract as a reviewer of modeling methods for the Ohio Depanment of Transponation, Rail Division.

FORECASTS Model Review andRevision Funded by Office of the Joint Chiefs of Staff, Department of Defense

].

1985 - present The FORECASTS model, an unclassified global forecasting model, is being reviewed and ;

substantially revised by a project team including staff of Resource Systems Group and i faculty of the Naval Postgraduate School. >

Adict Rdsum6 The Effects ofAcidRain Concentration on Structures: A Damage Function Approach Funded by U.S. EPA through the Cold Regions Research and Engineering Laboratory January - June 1984 This project developed statistical and simulation models to describe the effects of acid rain on structures. The objective was to quantify this component of the costs associated with increased acidity of precipitation.

Analysis ofAlternative Wood Transport Systems Funded by U.S. Forest Service, Washington, D.C.

October 1982 - September 1984 In the recent past, there have been significant increases in the use of wood as both a residential and commercial energy source. Both the delivered price of wood fuel and the impacts of this wood use depend heavily on the wood transpon system. This study reviewed and empirically evaluated existing systems in and effon to develop guidelines for ;

new systems.

Technical Review of the Oak Ridge National Laboratory's (ORNL) Highway Gasoline DemandModel Funded by U.S. DOE, Energy Information Administration (EIA) through Science Management Corporation, Landover, MD November 1981 - March 1982 The ORNL model was the only one available which forecasted state-level gasoline demand.

The U.S. DOE, EIA requested this study at the Thayer School to Review the technical adequacy of the model and to suggest ways in which the model structure could be improved. The model sectors (new car demand, auto scrappage, gasoline demand, fuel efficiency, and new technology) were reviewed in detail and a repon suggesting several major structural changes was prepared.

Integrated Paratransit Planning Methods Funded by U.S. DOT, Washington, D.C.

October 1981 April 1984 With George A.Kocur This project developed a microcomputer-based planning system for small-city and speciahzed transit systems. Extended performance, cost and ridership models were developed in a structured, goal-oriented computer environment.

Strategic Assessment ofFossil Energy Options, Phase 11 Funded by U.S. DOE, Washington, D.C.

September 1981 - April 1983 With George A.Kocur This continuation of previous work at Thayer School for the DOE Fossil Energy Division included an analysis of specific current and future fossil energy and uses. Pan of the study was an evaluation of the effects of transponation on the long-tenn competitiveness of U.S.

coal in the world market. Costs and capacities of U.S. inland transportation systems, transshipment (pon) facilities, ocean transponation, and foreign inland transportanon systems were compiled to support the analysis. A second pan of the study was an analysis i of the effects of changing emde oil prices on the domestic diesel engine market. For this i

Adler R6 sum 6 l analysis, a model was developed which represented the effects of changes in diesel fuel price on diesel engine demand.

Performance Standards and Fare Policiesfor Urban Mass Transit Systems Funded by U.S. DOT, Washington, D.C.

July 1980- July 1983 With George A.Kocur This study developed optimal bus transit system designs for a series of objectives which might be held by a local transit system. The results were given as closed-form expassions into which local parameter values were substituted to find the best system for a given city and given objectives. The equations gave desired average values of route spacing, stop spacing, headway, route length and fam levels. Warrants for express service and articulated buses were also developed. The second phase of the project developed manuals for transit operators and planness to implement these designs and warrants.

Direct Utility Assessment Funded by U.S. DOT, Washington, D.C.

August 1980 - November 1982 This project developed techniques and a handbook for using experimental designs and survey techniques to analyze market responses to transportanon innovations. The work combined aspects of discrete probabilistic choice models and experimental psychology to analyze situations that am difficult to treat with either alone.

Transportation Energy Policy Study Funded by the Solar Energy Research Institute Contract No. AB-0-4301-1 Maich 1980 - September 1980 This study used a transportation / energy forecasting model (ENTRANS) developed at the Thayer School of Engineering to cietermine the long-r age effectiveness of various energy conservation measures. The effort's primary client was SERI's Transportation Task Force, which was responsible for developing quantitative estimates of gasoline consumption levels as input to the Institute's general solar conservation study. In addition, the results were useful to, and directly used by, the several other federal a gencies involved in tmnsportation and energy policy-making. The specific policies analyzed included: auto fuel efficiency mandates, gasoline taxes, auto excise taxes, and public transit investment.

Interactions between National Energy Supply and Transportation-Related Energy Consumption Funded by U.S. DOT Contract DOT-RC-82003 September 1978 - January 1980 Rapid depletion of domestic oil and gas reserves will have a major effect on the price and availability of energy over the next few decades. The transportation sector will not benefit significantly from a national shift tc, ese of coal and nuclear electrical generation because almost all of the transportation, use is currently in the form ofliquid fuel. The purpose of this msearch was to analyze trar'sportation policies, such as construction of public transit facilities and development of electric cars, in terms of their long-range impacts on the use of transportation services and on the consumption of energy. The effects of time delay on the implementation of transportation technologies were analyzed. Interactions between energy supplies and transportation-related energy use were incorporated in a computer model used !'

by the U.S. DOT in its policy analyses.

f, -,-_------_-._--_------------

I i<, '

Adler R6sumd - .

I Analysis ofProductivity and Efficiency ofRural Transit Systems

. Funded by U.S. DOT, Office of the Secretary, Contract DOT-OS-70079 September 1977_- April 1979 '

The availability of public transit systems in rural areas has generally ' declined over the past two decades. Major new federal programs, however, are beginning to provide money for support of existing or proposed rural transit systems. This research will determine how these systems can be designed so as to be maximally effective in providing transportation .

service. In working toward this objective, detailed case studies and statistical studies are being conducted to relate productivity and efficiency with different design factors. In addition, a general simulation modelis being developed to trace a transit systerr/s evolution over time, given different design strategies.

.i 1

l 1

i 4

i:

.['

RESUME COLIN J. HIGH ReseArch Associate Professor Principal Thayer School of Engineering Resource Systems Group Dartmouth College Route 5 Hanover, NH. 03755 South P.O. Box 1104 Norwich, VT 05055 DEGREES B.Sc. University of Bristol, England, 1965 Ph.D. University of Bristol, England, 1972

' EDUCATION 1962-1965l University of Bristol Special Honors Geography / Geology, Minor in Economics.

1965-1970

. University of Bristol, Ph.D. program in Geomorphology. ' Thesis:

Aspects of the Solution Erosion of Limestones with a Special Consideration of Lithological Factors.

POSITIONS EELD' 1979-Present Research Associate Professor of Engineering and Environmental Studies, Dartmouth College, Hanover, NH.

1973-1979 Assistant Professor of Geography, Columbia University, New York, New York.

1977 Research Associate in Engineering, Thayer School of Engineering, Dartmouth . College, Hanover, NH.

1972 Visiting Assistant' of Florida, Professor Gainesville, of Engineering Sc$ence, University Florida.

1968-1972 Lecturer in Geography', University of Ibadan, Nigeria.

-1967 Project Manager, Hydrology Project. Jamaica. Supported by UN Development Program (3 months).

RESEARCH PROJECTS 1985-88 Acid Deposition Monitoring on Mt. Moosilauke.

Funded by USEPA 1

Cooperative Agreement #CR 812438-01-0 $206,747 (1st year of 3 years)

I Principal Investigators: '

Ryan (Environmental Studies) James Hornig, Colin High, and Douglas 1985-86 Wood Energy Technology Conversion Assistance Program: Strategic Analysis and Program Evaluation Funded by the Coalition of Northeastern Governors and the U.S. doe. Contract $35,000 - (Thayer)

I'rincipal Investigator: Colin High

t 2 1

l

.1985 Test of The Mountain Cloud Water Collector Funded by New York Botanical Garden and the USEPA Grant: $3,000 (3 months)

Principal Investigators: Colin High and James Hornig 1985 Economic Analysis of the Impact of Wood Energy in The Northeast.

Funded by Coalition of Northeant Governors and the U.S. DOE Contract # DE-EF 05-830R21389 $59,993 (1 year)

Principal Investigator: Colin High l

1983-84 Evaluation of major long-term environmental trends and emerging problems in New England.

Funded by U.S. EPA Grant #CR-811187-01-0 $57,113 (1 year)

Principal Investigator: Colin High (complete) 1983-84 Wood Energy Forecasting Model for the United States Funded by U.S. Forest Service Grant #FP-83-0570 $45,634 (2 years)

Principal Investigator: Colin High (complete) 1983-84 Assessment of the Application of Remote Sensing Techniques in the Col-lection of Agricultural Statistics Funded by U.S. Department of Agriculture Grant #58-319-V-3-0444X $37,096 (1 year)

Principal Investigator: Colin High (complete) 1980-83 Analysis Support Service for Strategic Assessment and Special Studies.

Funded by U.S. Environmental Protection Agency.

Grant #CR808-12-01-4 $223,313 Principal Investigator: Colin High (complete) 1980-82 Assessment of the Environmental Impact of Biomass Energy Systems in New England.

Funded by Northeast Solar Energy Center (U.S. DOE)

Grant #79-9-007 $133,135 (2 years)

Principal Investigator: Colin High (complete) 1980-81 National Biomass Energy Program Environmental Assessment.

Funded by U.S. DOE / North East Solar Energy Center Crant #80-10-003 $21,246 Principal Investigators: Colin High and William Lockeretz, Tufts Uni-versity (complete) 1979 Environmental Assessment of the Impacts of Large Scale Mechanized Harvesting on Vermont's Forest.

Funded by State of Vermont.

l Grant #5766215 $10,000 Principal Investigator: Colin High (complete) l l 1978-80 Environmental Impact Assessment of a Wood-Fired Electric Power Plant

! in Maine.

l Funded by U.S. DOE I

3 q

l. Contract #PO 1899-4,'21-1899-10. . $350,000 Principal Investigators: Colin High and Charles Hewett (complete) 1977-78 Study of the Feasibility of' a Wood-Fired Electric Power Plant in

. Vermont Funded by.the State of Vermont.and the Depar ment of Commerce Grant # DOC 01-6-06159 $39,500 Principal. Investigators: Colin High and Charles Hewett (complete) 1977 Assessment of the Use of Satellite Remote Sensing Data for Agricul-tural'and Forest Land Use Modeling.

Funded by NASA .

$25,000 Principal Investigators: Colin High and Dennis Meadows (complete) 1975 Use of. Remote Sensing Techniques for Agro-Environmental Studies.

Funded by NASA grant to Columbia University Grant # $180,000 (3 years)

Principal Investigators: Colin High, Jerry Coiner and Sam Goward (complete)

.1973 Modeling the Effects of Nuclear . Power Plant Construction on the Crystal River Estuary Florida.

Funded by a grant from the Florida Power Corporation to the University.

of. Florida Principal Investigator: Howard T. Odum, University of Florida.

TEACHING.

Courses Taught: Present The Earth as an Ecosystem - an introduction to environmental science (ENVS 2) l Energy and Resources (ENVS/ GEOG 12)

Environmental Policy Formulation (ENVS 50)

Seminar on the Environmental Problems of Kenya'(Foreign Study - ENVS 84)

Environmental issues in East Africa.(Foreign Study - ENVS 40)

Resource Systems (ENGG 177 - graduate)

Courses Taught: Past Resource Policy (ENGG 178 - graduate)

Agricultural - Forestry Policy (ENGG 179 - graduate)

Physical Geography Geomorphology-Hydrology Statistical Methods in Geography Biogeography (graduate)

Models in Resource Management (graduate)

Other Teaching Activities:

Advisor to five MALS students Director of the Environmental Studies Kenya Foreign Studies Program 1983-84 Member of the Faculty Steering Committee of Environmental Studies Adjunct Member of the Faculty of the Department of Geography

4 Member of the faculty of the New England Studies Institute of the MALS program supported by a grant from the National Endowment for the Humanities Formerly advisor and exan.iner for 14 MS, D Phill and Ph.D. candidates in Geography and Geology at Columbia University Formerly a member of the Undergraduate Committee on Instruction and Student Petitions Committee at Columbia University Formerly member of the Admissions Committee for the Graduate Program in Geography at Columbia University PROFESSIONAL AND PUBLIC SERVICE ACTIVITIES 1985-86 Consultant to Town of Sherburne, Vermont on Environmental Trends and Growth.

1985-86 Consultant to New Hampshire Attorney General for Public Hearings en the Environmental Impact of the Construction and Operation of a 450 kv DC Electric Power Transmission Line from Commerford, NH to Massachusetts.

1985-86 Member Technical Advisory Board of the Coalition of Northeastern Governors' Regional Biomass Program.

1983-84 Chairman of the Steering Committee of the Council of North Eastern Governors Biomass Energy Project which advises the council on the ex-penditure of $1 million per year on biomass research and development in the northeast.

1982-83 Member of the Vermont Natural Resource Council Panel on the effects of the high voltage electric power line from Quebec to Comerford passing through northern Vermont. Prepared the VNRC position and response to the Draft Environmental Impact Statement.

1983 Advisor to EPA Region 1 in the preparation of the New England annual

" Environmental Management Report." This is the major planning and budget justification activity for EPA regions.

1982 Consultant to the New York State Commission on Tug Hill for planning forest management programs.

1982 Consultant to Evergreen Energy Corporation on the environmental effects of a wood to methanol conversion plant.

1981 Consultant to the New Hampshire Governors Energy Council Study on the f easibility of producing methanol f rom wood in the state.

1981 Reviewer for the NSF Science for Citizens Program proposals.

1981 Member of the Advisory Panel for the New York State Adirondack Park Agency Land Use and Forest Practice Review.

1981 Consultant for the U.S. Congressional Office of Technology Assessment 011 Shale Study.

i 1

5 l

1

-l 1980 Consultant for the U.S. Congressional Office of Technology Assessment Energy from Biological Processes Study. l

?

1980 Consultant for the Argonne National Laboratory Residential Wood Energy Program Plan.

1980 Consultant for the Northern Community Investment Corporation feasibil-ity study of a wood chip pelletization plant in New Hampshire.

1980 Reviewer for the Oak Ridge National Laboratory Biomass Environmental i Assessment Study.

f 1980 Consultant to the MITRE Corporation and the Tennessee Valley Authority Wood Energy Program.

4 1980 External Reviewer for the Antioch College Community Wood Energy Study.

1980 Expert witness for the city of Winooski at the Vermont Public Service Board hearing on a permit application by Burlington Electric Power Department to construct a wood-fired power plant in Burlington.

1979 Consultant to Mittelhauser Corporation for a wood energy health and safety study.

1978 Expert witness for the State of California Energy Research and Devel-opment Agency at public hearings on Pacific Gas and Electric. Co. plans to use renewable energy resources for power generation.

1977 Consultant to the Wheelabrator Frye Corporation for assistance in pre-paration of a major biomass energy development program proposal to the U.S. DOE.

PUBLICATIONS AND PRESENTED PAPERS (parcial list)

Forecasts of Industrial and Commercial Wood Energy Use 1985-2000 in preparation for Journal of Forestry Products Research.

1 Economic Impact of Wood Energy Use in the Northeastern States, in preparation l for Energy Policy.

Economic Impact of Wood Energy Use in the Northeastern States: Technical Report and Appendix. Coalition of Northeast Governors, Washington, DC., 1986 (in press).

Region I: Assessment of Environmental Trends and Emerging Problems in the New England Region. USEPA Report. Final publication pending.

Preliminary Projections of Industrial Wood Energy Consumption in the U.S.

(4 volumes - 1 for each Forest Service Region) U.S. Forest Service Report.

Final publication pending.

Environmental Assessment of the Department of Energy Biomass Energy Program with William Lockeretz and others. U.S. Department of Energy (1982).

J' 'a (/

6 Wood Energy in the United States. Annual Review of Energy, vol. 6, 1981, with Charles Hewett,'Forman Marshall, and Robin Wildermuth. 1 Improvements to the Micro-Erosion Meter. With S. Trudgill and F.K. Hanna. I British Geomorphological Research Group Technien1 Bulletin No. 29. pp. 3-17, 1981.

j.

The Wood Resource and Its Use for Energy in the United States. . Proceedings 1081 l International) Conference on Residential Solid Fuels, pp. 670-679. Portland, Oregon, 1981.

An Environmental Report on an Advanced System Demonstration for Utilization of Biomass as an Energy Source in Westbrook, Maine. With others. Divicion for Biomass Energy Systems, U.S. Dept. of Energy, 1980.

Environmental Aspects of Wood Energy Conversion. With Charles Hewett. Pro-ceedings of U.S. Department of Energy Environmental- Control Symposium, Washington, D.C., November 28-30, 1978.

The Construction and Operation of Small, Dispersed, Wood-fired Power Plants in New England. With Charles Hewett. Report prepared as testimony for the California State Energy Resources Conservation and Development Commission, under Contract #500-144, 1978, Evaluation of Locations for a 50-Mw Wood-fired Power Plant in Northern Vermont.

Report for Vermont State Energy Of fice, under Grant #01-6-01659 from the U.S.

Department of Commerce, Economic Development Administration, 1978.

The Feasibility of Operating a 50-Mw Wood-fired Power Plant in Northern Vermont: !

Wood Availability and Siting Potential. With Charles Hewett. Report prepared for the Ve rmont State Energy Office under Grant #01-6-01659 f rom the U.S. i Department of Commerce, Economic Development Administration, 1978.

s Application of Digital Analysis of MSS Data to Agro-Environmental Studies. With 1 J.C. Coiner and K.E. # ebb. Three reports of studies under NASA Grant NSC080, National Technical Infonnation Service, 1976, 1977 and 1978.

Envir amental Impacts of Harvesting Non-commercial Wood for Energy: Research Progra s. DSD #101, Resource Policy Center. Thayer School of Engineering, j Hanover, N.H. Paper presented at U.S. Department of Energy Workshop on research l

.needs to determine environmental and soil productivity effects of residue l removal from agronomic and silviculture crops. University of Illinois, Oct. i 1977.

Digital Processing , 7 SAT Data for the Survey of Rural Land Une in the West African Savannah. Wiih d. Wilson. Presented at the Association of American Geographers Conference, New York, April 1976.

Monitoring Tropical Land Cover Dynamics From Remotely Sensed Geograrhic Data.

With J.C. Coiner. Proceedings 23rd International Geographical Congrese; Moscow, 1976.

Ground Level Rain Cauge Catch at 1badan, Nigeria, Weather, 29 (10): pp. 384-386. With J.S. Oguntoyinbo (1974).

i x 7 e

,,, c W. , s

/. Rainfh!1,DroughtandFoodSupplyinSouth-westernNigeria, Savanna,2(2): pp. I L" 115--'320. With J.S. Oguntoyinbo and P. Richards 1973.

]

Land Evaluatien Studies with Special Reference to Nigeria, in K.M. Barbour, ed.,

Planning for Nigeria: A Geographical Approach (Ibadan, Nigeria: Ibadan University Press, 1972).

]

t

{ -

l' Anomalous Relationships between'7:eunder Wave-lengths and Passage Width for Three Irish Caves, discussion paper, Awerican Joernal of Science, vol. 169, pp. 494- 1 496 (1970). ,

l A Method for the D,irect Measurement of Erosion on Rock Surfaces," British Geo-morphological Reses:ch Group Technical Bulletin, No. 5, 24 pp. With F.K. Hanna (1970).

Planning and the Natural Environment, Nigerian Cpinion 6 (5-7): pp. 60-64.

With A. Faniran and J.S. Oguntoyinbo;(1970).

t

/

r 4

I r

.d I

u 1

l r

l l

i l

1-~ I

g RESUME-1 DENNIS.L. MEADOWS (Professor of Engineering.

. Resource' Policy Center Thayer' School of-Engineering Dartmouth College Hanover,'New Hampshire 03755 Born: June 7, 1942 EDUCATION B.A. Carleton College, 1964 - Magna Cum Laude

. Ph .'D . Massachusetts Institute of Technology, 1969 FIELDS System. Dynamics, Information and Control Systems, Besearch Management, Microcomputer-based tools for management training and decision support.

PROFESSIONAL EXPERIENCE

, 1969-72: Assistant Professor, A.P. Sloan School of Management, M.I.T.

o 1972-78 Associate Professor of' Engineering, Dartmouth College 1978- Professor of Engineering, Dartmouth College 1973 Senior Fellow, East-West Population Institute 1974- Consultant.on Systems Methodology to Office of Technology Assessment, General Accounting Office, Congressional House Committee on Energy and Commerce, Solar Energy Research Institute & other Federal i agencies . .

1978-79 Visiting Scientist, Norwegian National Council for Science &

Technology 1980-83 Member, Electric Power Research Institute Advisory Council j 1983-84 Director, Integrative & Special Studies, International Institute for

' Applied Systems Analysis, Laxenburg, Austria

l. '1984-' Associate Editor, Simulation &, Game _s, Technological _Forecas_t_ing and_

Social Change, System Dynamicss _Louroal n

'1985- U.S. Project Leader fer the Soviet-American Bilateral Agreement in Environmental Education 1986- President, International System Dynamics Society 1986 . Member, Governing Council, International Simulation and Game Association 110NORARY AND PROFESSIONAL SOCIETIES

' Phi Beta Kappa Council on Foreign Relations

. Society of the Sigma Xi The Society for Computer Simulation The-Club of Rome International Simulation & Gaming Assn.

The Cosmos Club North American Simulation & Gaming Assn.

p International System Dynamics Society AREAS OF INTEREST Developing the theory and application of computer simulation models for analysis of complex socio-economic systems. Management of research laboratories, systems sciences in East Europe. Design and use of micro-computer-based management' training games.

0 'p' .%

g. { -

d . ..

n Dennis.L< Meadows-

//.4 L24 . Resume r

....a u CURFDrriKESPARCH A

v Director of The'Balaton Group, a consortium of institutes in 22 countries Q j that focuses on policies to provide very high resource productivity..on a

. sustained. basis, Design 6r of microcomputer-based management trainins tools a used in 25 countries.

/ i t

[l PUBLICATIONS'(PaTtial Listing) '

ji

[ The Dynami_cs_c.,f_Commo_d,1,ty_ Produc_t,1,on, Cy,cJ es, ' Wright-Allen Press, . Camb ridge , MA ,

, ' y 1970. '

! a T1c? Lim tutto Crywt_h_, Universe Books, New York, 1972 (with D.H. Meadows, J.

gG Randers, W. Betr ms). e t

P "Tne Dynamics of Sc,11d Waste," with J. Randef.s, Technology Rcview, Cambridge,

. MA, March /Apri1L1972.

l,

>y "Toward a Scienca of Social Forecasting," Proc. of the htintisi Academy of.

-Sc_iences, Vol. 69 No. 12, December 1972.

Toward Global F J uf1'i'brium,- WriPht-A1Jen Press,' Cambridge, EA, l1973 (co-edited with D.H. Meadows).

, Dwaaier r_f Grogth in_ d Fini_t_e, World, Wright-Allen Press, Cambridge, MA, 1974, U!.L. Meadows et aL.). <-

"The Trsusition to ' Cpa.1.," with R. Naill and J. Stanley-Miller, Technology,

'uviev,; October /Nove'aber 1975. i "A Critiqud of the:Short-Term Perspectives Implicit in Most Resource Models,"

in. M_fy_e_ral MaterialOJo,deling, William Vogeley, ed. , Brookings Institution, 1975.' '

l "Iowards a. Forecasting Nodel of Energy Politics: International I Perspectives,"

' vith N. Chesceri and D.3. Ross, J. of Peace Science,' Spring 1976.

Alternatives to Growth - I, ed., Ballinger Pub]dshirg Ccmpany', Cambr:ldge, MA,

- 1977 f) .I

} m

=

'"A Physio-Psycho-Social Computer Model of the' Pathogenesis of Alcoholism," with U. Coluke <nd R. Landeen, in T}p Efology of Alcohol, ism, Vol 7, 1980.'

l u

'; Tea! Facts Hindering Utility Support for Solar . Energy," in Community _ Energy pelf-Reliance, Solar Energy Research Institute, US Govt. Printing Office, July 1980.

j,- '

[ W "A. Critique of the IIASA Energy Models," The Energy Journsi,.Vol. 2, No. 3 .'

MM.

/ .Dvhamics of Crowth in Solar Markets - A New Perspective on the Debate Over

~ ~

f ( '

Federa1_]n;c[eht tv3~,'~[1 th 01av'((e1Ts~a~,~~RTio~u~[c~i~[oH cy Centr.r, IEP7/T5T,'ifasiv~er,

,, NH, 1982.

" Die Grenzec des Vachstums - 17 Jahre Danach," Die Umschau, No. 4 February 17, 1984, i

I Dannis L. Meadows 3 Resume PUBLICATIONS (continued)

"STRATEGEM-2: A Microcomputer Based Operational Came on the Kondratiev Cycle,"

l IIASA, WP-84-60, Laxenburg, Austria, August 1984.

l User's Manual for STRATEGEM-1; a microcomputer-based management training game on energy-environment interactions, Resource Policy Center Working Paper DSD#509.

l l

December 1986 I l

l i

-7.

-]

w ,

-f- !

E l ,

ATTACHMENT 2 r

Revised SAPL Contention No. 31:

The evacuation time estimate report, as described in Volume 6 of NHRERP Rev. 2 does not meet the requirements of 10 CFR S 50.47(a)(1),

S 50.47(b)(10) and NUREG-0654 II.J.2, II.J.10 i, 10 h and 10 1, and Appendix 4 because it fails to account properly for the number of vehicles that would be evacuating the EPZ; relies in part upon unsupported' assumptions; relies in part t upon potentially biased input data; does not rely upon an extensive enough empirical base; relies upon traffic control personnel not shown to be available; does not appropriately account for travel impediments such as flooding, snow, fog and icing of roadways; does not account for the effect of driver disobedience on evacuation time estimates (ETEs); does not appropriately deal with topographical features; does not deal realistically with the transport of transit dependent persons; in some instances overestimates roadway capacity and, for all of these reasons, underestimates the amount of time it would take to evacuate the EPZ and its subparts (" Regions) under the various scenarios analyzed.

Appendix, Board's Memorandum and Order, May 18, 1987 at

p. 4.

Admitted Bases: '

1. This latest revision of the KLD Report now notes that 3,000 "through" vehicles will be in the EPZ highway network at the time of the order to evacuate (Vol. 6, pp. 2-27 and 10-3). j This estimate of 3,000 through vehicles at any 1 one time is unsupported since in 1985, traffic levels on I-95 alone in New Hampshire exceeded 99,0'00 vehicles per day, many of which were i through vehicles. The size of the Seabrook Station EPZ is roughly 200 square miles (Vol. 6,

p. 4-2). It is clear that the number of vehicles chosen by KLD significantly underestimates roadway demand. This error is particularly serious in the beach areas during i the summer season. The lack of appropriate l consideration of cars in the roadway system i contributes to the serious underestimate by KLD I of vehicles in the beach grea. Though the KLD i Report states that a total of 3000 cars were {

I i

1 i

)

!, . I:$ '

G' countediinfthe roadway from aerial-films in

.K

.Hampton' Beach,1there is no statement'as.to when those photographs wereLtaken 'ro by whomLthey 6

H were:taken (Vol'.7 6, p. 10-16). :It is not even

.. clear that KLD hasLfactored the: 3,000' estimate

!into?itsitime. estimate calculations since'the estimate'is_found in'a-sectionLeaptioned:

m '" Uncertainties."

. 4 2.. The KLD Report..continuesfto' rely..upon NRC estimates compiled in a report'by M, Kaltman'in .

February 1981 for estimate of vehiclessper

~

dwellingiat~ seasonal housing units-(Vol. 6,>

'p. 2-14), the count'of' overnight accommodation

, units-(Vol. 6, p._2-17), and numbers'of vehicles aticampgrounds (Vol. 6, p. 2-17), and numbers of vehicles-at campgrounds (Vol. 6, p. 2-20).

These-numbers.are'unlikely to have a:high

, reliability given the significant growth that has occurred in the EPZ area.over the past_5-

. years.

3.- The KLD Report has failed to make-any

provision.forLresident'and employee population.

as "

' growth in the area overthe_ plant's projected L11fespan.- ! Population redistribution in the' area.

could also' markedly affect: evacuation. time estimates (ETEs)'for various regions in the EPZ. At current rates'of growth,, resident populations w;111 increase over'70% and; employment will' increase to an even. greater degree.over'this time span. The result will be that the ETEs set'out in this report will

. rapidly be rendered obsolescent, even if one were to assume ~that they are currently correct.

4. The KLD Study' continues unrealistically I to assume that. traffic-management and control measures'are in effect'at the time the evacuation'is ordered (Vol. 6, p. 10-70 and Appendix 1). .This is a' wholly unrealistic assumption even if one were to assume that all local ~ communities.were intending to. implement the emergency plan. Seven towns did not participate in-the graded FEMA exercise on February 26, 1986. Under many plausible .

scenarios, state resources'and manpower, even if adequate, could not'be mobilized quickly enough

to~be in place in the non-participating or even

, the participating communities before it would be

= _ _ _____. -

necessary to order.an evacuation. The sensitivity calculation by KLD that an immediate General Emergency would extend ETE by only 20-30

l. minutes is non-persuasive and the assumptions L employed.in doing that sensitivity study are not

! carefully elucidated.

5. The KLD " planning-basis" accident scenario assumes that there are two temporally ,

displaced evacuation stages, i.e., the rest of !

the EPZ starts to evacuate 25 minutes after those in the beach areas (Vol. 6, p. 4-17).

This is not.a realistic assumption because many people will choose to leave the areas when they see the beach people going by, particularly given the heightened awareness of radiation health effects resultant upon the extensive news coverage of the Chernobyl accident. There is no i empirical basis shown for the KLD Report's 25 percent spontaneous evacuation rate among those within the EpZ but not ordered to' evacuate (Vol.

6, p. 10-3). The KLD Report fails to account for.those beyond the EPZ who may spontaneously evacuate, taking up roadway capacity beyond the zone and thereby impeding evacuation progress out of the EPZ.

6. The KLD Study relies upon potentially biased input data in that the telephone survey of " heads of households" provides a substantial portion of the data used. This survey involved calls to over 10,000 households, yet resulted in only 1,300 completed responses (Vol. 6, Appendices F and G). Given the. low response rate and the fact that no efforts were made to validate the responses, a large non-response bias exists in the completed data sample. There is, therefore, no reasonable basis for assuming that the notification times, estimates of times to commence evacuation trips, estimates of average person occupancy of vehicles evacuating ,

the EPZ or other data derived from the survey 1 are accurate (Vol. 6, p. 2-3, Exhibit 2-1 and Chapter 4).

7. Further, the " Time to Travel Home" data derived from the telephone survey (Vol. 6, p.

4-10) is of limited, if any, utility. Question

9 of the telephone survey asked: Approximately how long does it take Commuter #( ) to travel home from work or college?" This question as

Lframedfelicitsta response ab'out how long-it

' circumstances.

takes.the commuter to return home under normal Under the circumstances of~an evacuation,

'home-partly against however, commuters will be' returning the direction-of evacuating i:L traffic, through intersections.with cones

' blocking-' desired turning movements, and partly

.with queues. the' flow of evacuating traffic in massive y The:effect of almost 95-percent.of-the commuting population attempting to return home withinl30 minutes of each;other (Vol. 6, p. 4-9) would evacuation be.a massive rush hour even without an in progress. The assumption that-commuting workers can return home in their normal time frames defies common sense and.is -

insupportable from.an analytic standpoint.

'8. The KLD-Report still relies upon Traffic.

Guides for Traffic Control Posts (TCP) and Access Control Posts (ACP). 181 local and 10 interstate traffic. guides are needed for the TCP, 118 of whom are needed.in New' Hampshire

-(Vol. 6, Table 8-6). . An additional 130 personnel are needed at the ACP, 28 of whom are needed in New Hampshire (Vol. 6, Table 9-4),

bringing the-total New. Hampshire traffic .

personnel requirement to 146. The NHRERP still does not' support a finding that these personnel will be available in adequate numbers.

9. The KLD Report still continues to assume that all roads will remain passable during evacuation (Vol. 6, p.'10-70). This assumption is insupportable as it denies the realiatic potential for. vehicle problems either due to mechanical temperature.

malfunctions or extremes of The KLD Study acknowledges that

-temperatures in the EPZ range "from well below l

zero (F) in the winter'to as high as 100 degrees (F) in the summer" (Vol. 6, p. 1-8) and then ignores the implications of the statement for vehicle reliability. The KLD Study assumes that

.all vehicles with problems can be pushed to the side of the roadway. This ignores the bridges and other choke points within the EPZ, such as

.the entry onto Route 51 (a major evacuation route from the beach) which could be totally obstructed by one failed vehicle. In those situtations no easy solution could be effected since there is no place to push the vehicle.

Even if a vehicle is pushed onto the shoulder, l

1

- _ _ _ . - _ - i

m 3

~

=the-Highway Capacity Manual' estimates that capacity:1s reduced by one-third because the roadway's' perceived width is reduced. The KLD.

-planLunrealistically' assumes that no reductions

~in: capacity 7cr: increases in trave 1' times will result from these'incidences of vehicle failure. The KLD_ plan recommends stationery

. placement?of tow trucks at locations specified in Table'12-l'. However,'the NHRERP does not

reflect: this' specific assignment of tow- trucks to specific locations.. Further, the KLD Report gives no. estimate of how long it;would take.a tow vehicle to(respond.to an incident _and then

. return to'its assigned location. Only 2.of-the recommended tow truck locations.are:within the EPZ in New Hampshire (Vol. 6,-Table 12-1).

Vehicles could also be disabled by. exhaustion

.of fuel supply or accident. -The KLD Report assumes.that most accidents will involve vehicles. traveling atLlow speeds 7and that thereforeithey,will not result in vehicle disablement (Vol. 6,'p.;12-3). .However,-the KLD Report.doesinot assume that-all traffic flow is

' low: speed in thatcbuses are assumed to travel from 40 to 50 MPH (Vol. 6, p. 11-20). The KLD Report"still?does not appropriately account for flooding, excessive snow, fogfand icing of roadways. . It now makes only passing mention of

. fog (Vo1~. 6, p. 3-11) .and indicates that the capacity reductions for snow'and rain.are-responsive to the problem. The-KLD estimates of w ~ capacity reduction for rain are, as SAPL stated before, too optimistic. This new version has.

changed the capacity-reduction for snow to 25%

from the 30% stated earlier,'a move in the wrong direction. The 1985 Highway capacity Manual cites detailed studies which show capacity reductions of 8% for a trace-snowfall plus 2.8%

for each 0.01 in./hr. water-equivalent snowfall. For a snowstorm accumulating 5" of snow over 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , the corresponding capacity.

I: reduction would'be over 40%; substantially more than assumed in KLD's analysis. Flooding could render a section of roadway wholly impassible.

10. [ Basis denied by Board Order of May 18, 1987)

p 'T "1 11... The-KLD Report.now does.contain maps

'includingLtopographical features, but'~the time

' estimate study does'not. account 1for these features other than to make brief mention of

them (Vol .: 6,1p. 1-5). The effect~on time estimates of the more hilly topography west of.

I-95'and the'effect of the choke points at bridges 1over: rivers and streams have been-ignored.

12. The mobilization time for buses has been.

modified in this Revision 2 version of the KLD-Report. .ItLis now' claimed on the basis of a telephone survey'of the. organizations which own and operate the buses that 50% of available buses: (as. opposed to the. earlier 62%).can be' mobilized within one hour of notification, and another-30%Jwithin the second hour, with the remainder.following in the third hour-(Vol. 6,

'p.111-19). The survey instrument and.the data on the replies are not included.in-the appendices of the report.

During off busi' ness hours, in particular, the

~

response rate could'not reasonably be expected

, -to be anywhere near this' favorable.

13. The revised KLD Report.now, computes the numberLof persons within the EPZ having no-vehicles available.and requiring transit services at-2,249, or'2.5%.of the:91,601.

. population in the 17 towns in New Hampshire.

Again, KLD has moved in the wrong direction in reducing'the prior estimate of 3%. The basis of KLD's calculation was the telephone survey, the L problems with which were discussed at 6. above.

This estimate is now less credible.than it was heretofore. In'Section VI of the "New Hampshire Response Actions to RAC~ Review of State and Local Radiological Emergency Response Plans -

August 1986," (hereinafter " Reply to RAC"), the RAC commented that the vehicle ownership data should be compared to census data. KLD responded that though the 1980 census data showed~that 5.5% of all households have no car l available, a 2.65% estimate is reasonable "in light of the expansion of car ownership during the intervening years." (Reply to RAC., p. 3).

The RAC rebutted this position with national. car i ownership statistics and KLD came back with the I

reply that only site-specific data are relevant 6-i 1

1

l.

t .

i for planning purposes (Reply RAC, pp. 4-5).

KLD's reply is. unpersuasive oecause of the

.c problems noted with regard to.the telephone l - survey . at: 6. above. -Unreliable site. specific

- data are not better'than the national data. KLD 3 '

attempts to shore up its argument by. referring' to-the NHCDA survey, which arrived at a similar

, - number. The allegedly corroborative data has its own reliability problems. SAPL holds that the number of.those requiring transport RJ >

assistance 1has been' seriously underestimated.

14. The revised KLD. report now estimates the j

, time for loading passengers at special facilities at 45-minutes (Vol. 6,Jr. 11-21).

They assume that the average elderly or disabled person'can board a bus in a 15 second mean headway. SAPL still finds this an

. unrealistically short period of time for loading special' facility populations along with.their

- necessary personal effects and medications. The ,

estimate ofitime for loading non-ambulatory persons, previously 0.67 hours7.75463e-4 days 0.0186 hours 1.107804e-4 weeks 2.54935e-5 months , appears to have been.omitted from this revised KLD Report.

There is still no estimate of'the number of non-ambulatory. persons outside of special facilities.

15. Though the revised KLD Report states that substantial detail on roadway geometrics was collected (Vol. 6, p. 1-10), the. rural roads were classified-into only 4 crude groups (Vol.

6, p. 3-7 and 3-8). The detailed data collected should confirm that all sections of each roadway included in a given class have minimum widths greater than or equal to those assumed. This i has not been demonstrated.

16 [ Basis denied by Board Order of May 18, 19871

17. The calculation increasing the number of people by'6% because the average vehicle is out of' service 6% of the time is not correct. A proper calculation would increase the number of permanent residents needing transit by more than 6% (Vol. 6, pp. 11-8 and 11-9) based on data in Figures 2-2 and 2-3.

18. The simulation model employed by KLD appears to have some serious defects:

P~5 9

~(a) It.is unclear ~how traffic

. control information.is handled in-the actual simulation'and

' produced-the'results'in 1 Appendices I and.N. As an example,-Appendix N seems to imply that 1,500ccars can enter-node.#11(Vol. 6, p. I-49)'from each of three-directions.

F,.

-To resolve. questions, a sample derivation of link capacities should be included for one: simple and one more complex link.- A-sample of actual flow at a' crowded intersection,. showing-all input's,' outputs and queues shouldLalso be included. .

-(b) Loading procedures are.not described in much detail.' The full loading results at one major loading point should be included.

.(c) It-appears that a .

l substantial amount of passing has been assumed since a factor of fd = 0.75 x (0.90.)'= 0.675 is.used to get-one;way from1two

~

way capacity.

(d) Appendix I shows . light traffic on many roads. It is not clear how, if-at all, these light traffic patterns have

'been treated in the simulation ,

model.- '

19. The estimate of 2.6 people per vehicle {

for permanent residents is unrealistic, H particularly for the first nour when people will be returning home or picking up family members (Vol. 6, p. 2-5). The. data from the actual ] 1 counts of vehicle occupancy collected in August i 1985.and July 4 weekend in 1986 do not support this estimate (Vol. 6, pp. 4-6 and 4-8). 1 q

1 I

i 1

-____-_.____.___.m _ _

Nl s

~

20.' The KLD' Report lacks-a sufficient!

" empirical base for computing the transient

'populat'ionJin'the.EPZ. KLD should have taken-E . extensive aerial photographs of the area during the height-of the beachJseason. The reliance o upon' indirect' inferences from_ beach blanket

' space and parkingJspaces:in indefensible when the real picture could have been taken in a

, systematic and thorough; fashion.

For.all of the above-stated reasons,.the KLD Report'in, Volume 6 of NHRERP Rev.'2 fails to

. provide a sufficient basis for a. finding of reasonable assurance that1the public can and

'will be protected in the event of'a radiological emergency.

Seacoast Anti-Pollution League's Contentions'on Revision 2 of.

the-New Hampshire Radiological Response Plan, Nov. 26,-1986, at pp.

7-17 (Revised. Contention.31), as admitted per Board

Memorandum and. order.of May 18, 1987, at 44..

The KLD study has overestimated the capacity.of certainLroads and intersections. For example, Route 1A N/S is classified as a " Medium" design road-(See KLD. Progress Report No. 1, ESTIMATION OF HIGHWAY CAPACITY, P. 46). Route 1A N/S is in some places very narrow, has a. steep grade.along at lease one;section and winds along certain~ sections. It has at at least two points almost right= angle turns. It should'be treated as a low design road. The traffic from the beach area of Hampton is to

' get off:the beach by turning left on either

-Highland Ave, or7 Church St. and then traveling west ~ bound on Rt 51. Alternate-routes for each beach population all involve travel north on Route 1A with left turns at either 101C, 101D, South Road or Washington ,I Road. The capacities of Highland Ave, and Church St. will quickly be overwhelmed so j that people will need to go north on lA in large numbers.

1 The overestimate of the capacity of' Route 1A therefore can have very serious implications'for accuracy of the ETE. j

{

-Seacoast Anhi-Pollution League's Fourth Supplemental Petition

. For' Leave To. Intervene, dated May 5, 1986 (contention 31), at

. p.

1987-at11, as 12.admitted per Board Memorandum and Order of July 16, Am%A mmm_m_a__.___m- _ . - _ - . _ -- -._ -

SAPL Contention 34:

The New Hampshire State and local plans do not meet.the requirement that there be maps showing the' population distribution around the facility as required at NUREG-0654 J.10.b and Appendix

4. Therefore, there is no reasonable assurance that adequate protective measures can and will be taken pursuant to 10 CPR S50.47(a)(1) and S50.47(b)(10).

I Appendix, Board's Memorandum and Order, May 18, 1987 Admitted Bases:

A similar contention to this was originally i submitted as SAPL Contention No. 6, which SAPL j withdrew. Subsequently, however, the NHRERP !

Rev. 2 has appeared and it has very

)

significantly and erroneously reduced the peak .

populations in the EPZ from those that appeared in the November 1985 NHRERP upon which contentions were initially filed. The November 1985 NHRERP stated the peak population for New Hampshire at 191,849. Rev. 2 has reduced this I peak number to 142,529 for a summer weekend and 142, 929 for summer midweek. The larger of these two new figures represents a reduction of 48,920 people from the original estimate.1/

In NUREG-0895, the Final Environmental Statement for Seabrook Station dated December 1982, the NRC Staff estimated the resident population for 1983 within 16 km (10 miles) of the site at 99,900 and estimated the 1983 peak transient population within the same area at 130,998, which totals to 230,898. These .

figures, again, were the estimates for 1983.

Table 1 in the Rev. 2 plans show the total EPZ peak population for 1986 at summer midweek at 1/

Note that there inconsistencies among the local plans for the summer midweek population for Hampton. Table 1 in the Hampton plan states the number at 34,337, while the East Kingston and Stratham plans, for example, state the number at 31,337.

ig 0&

I

.c

'229,726,' less than thefNRC. estimate of peak

, population for.1983, and the peak summer weekend population at- 231,292, an' increase of.only 394 l'

. people. .'This is patently and plainly absurd.

Rockingham County,'N.H.-is'one of the fastest growing areas in the-entire country. 1980 census figures.showed the Rockingham County.

permanent: population at- 190,345. 'The estimate of'Rockingham County' population for 1985 from the.New Hampshire: Office of State Planning is 214,510. .Therefore,' the average annual growth k rate for Rockingham County as a'whole is 2.5%.

The' area'along the' coast,.where the EPZ towns are' located,'can be expected.to have grown at a rate faster than the county wide average.

Therefore, one must reasonably assume that the EPZ population has increased by a minimum of 1 16,000 people-above-the 1983-NRC estimate of 230,898.

SAPL finds particularly: objectionable'the treatment of.the. population for'the' Town of Hampton in the Rev. 2 plans. The November'1985 version of'the'NHRERP stated the peak seasonal population at 110,000. The Rev. 2' plans state the peak' summer weekend population at 36,635, a reduction of'73,365. .NUREG-0654 states that

" Estimates of transient populations shall be developed using local data such as peak tourist

-volumes and employment data for large factories." .(NUREG-0654, Appendix 4, p. 4-3.)'

The Hampton selectmen. notified the. State by .

. letter of October 29, 1985 that the peak estimate of-110,000 was'"still lower-than traffic counts and local business figures indicate." (See Attachment A.) Therefore, the reductions of population in the beach area

-particularly and.the EPZ as a entirety are wholly unwarranted. This leads to the ineluctable conclusions that the population maps in the-plans are not accurate and that therefore adequate protective measures will not be provided for the health and safety of the public because the planning provisions in the NHRERP do not provide for the realistic demands that will be placed on emergency response resources, in clear violation of 10 CFR S 50.47(a)(1).

Seacoast Anti-Pollution League's Contentions on Revision 2 of the New Hampshire Radiological Response plan, Nov. 26, 1986 (contention 34), at pp. 30-32, admitted per Board Memorandum and Order of May 18, 1987 at 46. i 1.

L. - . __m_________.i_____'. _ _ . _ . _

l.

I q' SAPL Contention 37:

The NHRERP Rev. 2 fails to provide reasonable assurance of adequate public protection because an adequate number of emergency vehicles are not provided for in'the plans and further there is no assurance that effective use of these vehicles will be possible in view of a potential outgoing flow of evacuating traffic and a significant lack of drivers. Therefore, these plans do not meet the requirements of 10 CPR ,

l 550.47(a)(1), S50.47(b)(3), S50.47(b)(10) and NUREG-0654 II.J.10.g. and II.J.10.k.

Appendix, Board's Memorandum and Order, May 18, 1987 at 1 p.-5.

Admitted Bases:

NUREG-0654 II.J.10.g., which provides a further explication of the planning standard at 10 CFR 550.47(b)(10), states that there must be means of relocation of the population in the emergency planning zone. The planning standard at 10 CFR 550.47(b)(3) requires that there be arrangements for regnesting and " effectively using assistance resources." (emphasis added.) Neither of these requirements are met because the NHRERP Rev. 2 fails to provide for adequate transportation resources for both those numbers of evacuees in institutions (schools, day care centers, hospitals, nursing homes) and those not in possession of an indepe~ndent means of transport.

The NHRERP Rev. 2 has reduced the numbe buses to be provided under plans to 482,l/r of down from a prior 614 planned-for in November 1985 version of the NHRERP. The Town of Exeter and the City of Portsmouth are the municipalities most drastically affected by these reductions. The projected need for buses for the City of Portsmouth is 89 buses less than 1/

Based on numbers taken from the local plans. See, for example, Greenland RERP at IV-22, Kingston RERP at IV-12a, and Brentwook RERP at IV-21a).

, a

, 1 f

wasjprovidedgin.the'Nov.n1985'Portsmouth RERP.

' ' .This situationfbecomes'particularly serious f '

because.-the KLD Study shows traffic congestion

, Ein:theLCity-~.of Portsmouth throughout the evacuation process.- (Vol.: 6 at-pp. 10-21 to 10-27 and pp. 10-29.to 10-34.) If additional'

" transportation resources must be requested from-

.the State,'there could be a significant degree o .of1 difficulty in'getting buses through traffic

. to theJareas where they are needed. Further,

' PortsmouthLis the location of one of the transportation staging areas (at OMNE-Mall). The i l-Town'of Seabrook and Hampton Falls'are among the towns.to be se~rved by buses.from that staging area (vol. 4, p. I-10). Portsmouth is'in the J

northmost section of the New Hampshire. portion E of the EPZ. .Hampton: Falls and Seabrook are at

'the southernmost section ofcthe New Hampshire portion of the EPZ.

' Thisiprovides'a good-example of the fact that

.in.many instances, buses;will need to travel

'throughhvery congested traf fic' conditions over a considerable-distance before they can get.to a p.:nt of. effective use. Given the' constraints of personnel and resources faced by the local and'N.H. State Police, it is-not likely that adequate traffic control measures will besin effect. There will therefore be nothing to

' constrain evacuees from attempting to drive outbound on those lanes that are'needed for effective-progression inbound.of emergency vehicles. TheJN.~H.-State Police Troop-A only has 35 troopers, not an adequate number to cover-the.118,traffice conttol posts and 28. access control posts in New Hampshire. ('EMA, Final Exercise Assessment, 6/2/86'at p. 46 and Vol. 6, Table 8-6 and 9-4.) In Hampton Falls, for example, there is only one full time and four part, time police officers. Police.Chiaf

-Christie has stated that he is unable to fulfill the duties designated to him in the plan due to lack ~of manpower and equipment. The Town ef Hampton-has'only 24 full time officers in the

-police force. In addition to traffic control, these officers must provide. security at various locations and perform back-up alerting functions in addition to their usual functions.

Further, there is no assurance that there .

I will be drivers for even the inadequate numbers of. buses and vans provided by the NHRERP, 1

Rev.2. The letters-of agreement in Volume 5 show a total-of.574 buses and 510 drivers. The 64 driver shortfall for buses alone, not to mention the vans, does not really show the true magnitude of.the problem since the excess availability of a driver or drivers to one bus company does not do anything to alleviate a shortage of drivers at another company. FEMA stated that " bus-and-driver pairs" should be tallied to arrive at the maximum availability of emergency vehicles. (FEMA, Final Exercise Assesment, 6/2/86, p. 39). The letters of agreement in Volume 5 of the plans only demonstrate a total of 445 bus-and-driver pairs. As stated above, the local plans alone show a need for 482 buses, though the State plan shows that only 444 buses are needed for those communities Because the State of New Hampshire has erroneously reduced its estimate of the number of buses that will be needed by reducing ,

i population estimates and by reducing the l percentage of non-auto owning individuals to be provided for (see SAPL Contentions Nos. 34 and 18), because there are insufficient emergency vehicle drivers and because adequate traffic control measures cannot be carried out due to personal shortages, there is not reasonable assurance that adequate assistance resources will be provided or that those resources that are available can be used effectively.

(Those bases of Contention 37 that rely on " comparisons between the November 1985 version and Revision 2 of the NHRERP" are omitted per Board order of May 18, 1987.]

Seacoast Anti-Pollution League's Contentions on Revision 2 of the New Hampshire Radiological Response Plan, Nov. 26, 1986 (Contention 37), at pp. 35-38, as limited per Board order of May 18, 1987, at 50.

TOH Revised Contention III to Revision 2:

The Evacuation Time Estimate Study (ETE) prepared by KLD Associates, Inc., Revision 2, Volume 6, is based upon inaccurate and biased factual data and unreasonable or misleading assumptions, fails to comply with NRC regulations, and fails to provide reasonable assurance'that adequate protective measures can and will be taken, or that I adequate facilities, equipment, or personnel will be provided to the Town of Hampton, in the event of radiological emergency. 10 CFR S 50.47(a)(1), (b)(1)(10)

NUREG-0654, App. 4.

Appendix, Board's Order & Memorandum, May 18, 1987 Admitted Bases:

The KLD ETE is based upon the following inaccurate, unreasonable, or misleading facts or assumptions:

1. The KLD ETE unreasonably estimates vehicle counts within the EPZ, including beach areas, utilizing data obtained on only two weekends, and the intervening work week, in August, 1985. KLD Progress Report #1 l (hereinafter KLD #1) Appendix E-13. KLD )

relies upon these limited vehicle counts as part of "the basis for computer analysis of an Evacuation Plan and computation of ETE."

KLD #1, pp. 5,6. KLD concedes, however, that this traffic data was gathered during a period of " occasional rain," KLD #1, p. 7, "this period of time was not particularly appealing to beachgoers," KLD #1, Appendix E-13, "The data will not reflect peak conditions" . . . and there is "some uncertainty" on the accuracy of the data. On its face, therefore, the KLD ETE admits to an inadequate factual base to provide reasonable projections for traffic counts and movements q during an evacuation within the EPZ, and f particularly the beach areas. Additionally, since even this limited data was obtained by ,

KLD during poor beach weather, it must be assumed that KLD's vehicle counts, and therefore ETE projections, are unreasonably I low. !

l

2. The KLD ETE unreasonably relies upon a 1 telephone survey to estimate the time

^

JI::, '

7

( -s

~

require'd for, notification'of anLemergency, elapsed. times.to commence evacuation. trips, and the: total population to be' evacuated from the.EPZ. .KLD #1, p.- 7, KLD #2, p. 9. Those-persons surveyed constitute less than one percent of the individuals residing within the EPZ, KLD #1, Appendix.F-6, there is absolutely no showing by KLD that this minimal percentage of residents is in any way representative of the EPZ population as a whole, and.ltherefore,the telephone survey.

represents an inadequate factual base from which to make these ETE projections. For example, while KLD concedes that "we know of no. survey.which has accumulated empirical

~1nformation describing the rate at which notification information is received," KLD

2, p. 7, it nevertheless baldly claims that the telephone survey can provide "a reasonable estimate of a notification time frame." KLD #2, p. 7. The Town suggests,1 however, that limited information obtained by ,

telephone from an apparently nonrepre-

, sentative segment of the EPZ population is wholly inadequate to make these significant ETE projections. Based upon the admitted deficiencies in.its data base, therefore, the KLD ETE necessarily fails to provide reasonable assurance on the accuracy of.these ETE estimates.

3. The KLD ETE computes the number of vehicles to be evacuated from the beach areas merely by counting parking spaces and parking capacity. KLD #1, p. 15, 20. The KLD ETE therefore fails to account for the virtual bumper to bumper. traffic that routinely, and continually, travels through the beach areas during the summer. These vehicles in transit represent not only a significant additional number of vehicles to be evacuated, but also present a substantial impediment to all parked vehicles attempting to leave the EPZ.

See also KLD #1, Appendix E-4, 5; KLD #2, p.

9.

4. The KLD ETE erroneously assumes that local officials, including police and fire department personnel, will be available to implement the State evacuation plan. KLD #2,

-_-_i_L_-____- .___-

t'

~

p. 40. .Since_ the Town of Hampton'has stated Lit: does :not intend to provide this assistance k or' implement;the State RERP, the KLD ETE .

' fails to/ provide: reasonable assurance that adequate ~ personnel;are available'to implement.

h the1 evacuation' plan. Additionally, even o ' assuming that the-State'could timely provide an equivalent number of State personne1Lto

. fulfill these local functions, the KLD ETE i

recognizes the local personnel are uniquely

. qualified:to determine potential traffic 1

problems and bottlenecks, which may not be ;

readily apparent to State personnel unfamiliar with the local area. KLD #2',

p'. 40.

l S. The KLD ETE unreasonably assumes that i 151 " traffic guides" will be available to implement traffic control. procedures during an evacuation, including 25 for'the Town of !

Hampton.- KLD #4, p. 11. 'The KLD>ETE wholly I

fails,:however,.to' demonstrate the availability of these substantial numbers of trained' traffic personnel. Additionally,

'since State Police Troop A'has only.31- ,

troopers available for evacuation traffic control'throughout the entire EPZ, it is unreasonable to expect that the State can adequately and promptly supplement these a personnel deficiencies, particularly in view of the substantial additional duties imposed on. Troop'A for overall traffic surveillance, KLD #7, p. 28, and as specified in'the State Compensatory Plan.: See Compensatory Plan, Troop A New Hampshire State Police, Emergency Response Procedures, p. 2. The KLD ETE further unreasonably assumes the availability of an additional 27 New

. Hampshire " traffic guides" to regulate access control posts on the perimeter of the EPZ to restrict traffic entrance into the EPZ during an evacuation;n. KLD #6, p. 13. As set forth above, there is no showing that in fact l these trained personnel will be available to perform these specified duties.

6. The KLD ETE unreasonably assumes that adequate equipment and personnel will be available to plow roads and driveways, and to assure that evacuation routes remain 1

s

passable, iffevacuation is requ' ired during a

' snowstorm.- KLD #2, p. 19, 24. 'For example,.

h KLD incorrectly assumes that the time to-plow

- theidriveways during an evacuation is identical to.the' time required for snow clearance under non-emergency conditions.

The.ETE therefore. unreasonably-fails to-account for evacuation traffic congestion which must impede or prohibit a plow truck

. from reaching certain homes on roads, and unreasonably fails to consider-that a substantial number of those private individuals performing snowplow services may elect to promptly evacuate,the EPZ rather than complete: their routes.

7. The KLD ETE unreasonably assumes that

- buses will encounter "little impedance" when entering the EPZ to evacuate schools and those without private vehicles. KLD #7,

p. 17. This assumption is insupportable.

For example, the State RERP provides that the Timberlane Bus Company of Salem, New R Hampshi're shall' provide 35 buses to evacuate the Town of Hampton during a radiological emergency. These Timberlane buses, however, would be required to maneuver through thousands of evacuating. vehicles headed for the " host" communiti2s of Manchester and

~

Salem. KLD #4, Appendix J. It can only be

. reasonably anticipated that a substantial number'of these buses would be greatly delayed, if not prohibited, from reaching the EPZ against the evacuation traffic' flow. The KLD:ETE further unreasonably assumes that i evacuation buses traveling to the EPZ could travel 40 miles per hour on "at-grade primary highways," such as Route 1, and 50 miles per hour on access controlled roads. KLD #7,

p. 17. Anyone familiar with the routine bumper to bumper traffic on Route 1 during the' year, and particularly the summer months, however, would recognize these estimates as wholly unrealistic. While recognizing that i buses and vans evacuating special facilities i "will be embedded within the overall traffic streams evacuating the EPZ," KLD #7, p. 19, the ETE further unreasonably fails to account for the additional and substantial impact of these emergency vehicles, often traveling

- a i

1 ,

against1the. flow of. traffic,'in delaying the

^

overa11' evacuation of vehicles from the EPZ.

KLD;#2,,p.:9. :Further the KLD ETE unreasonably' calculates the time within which buses may travel evacuation routes to pick up.

passengers by assuming that all buses will travel'with the flow.of evacuating traffic.

KLD.#7, p. 18. It must be assumed, however, that many'of these buses will be required, albeit unsuccessfully, to travel against the flow of traffic to reach designated pickup locations. The ETE's calculations that buses may therefor be expected to travel through evacuation traffic and to reach and load passengers'at specia1' facilities within'40 minutes is plainly unrealistic. KLD #7,

p. 18.

8. The'KLD ETE unreasonably relies upon

, inadequate date to compute the number of persons to be evacuated from the EPZ in the event.of radiological emergency. First,~KLD

. computes.overall' population figures based upon a " compromise estimate" of 2.8 persons per vehicle, although KLD' concedes that it lacks " definitive data" on this issue. KLD-

2, p. 9. Second, as previously discussed, KLD relies upon'a telephone survey of less than one percent of EPZ residents, without any determination that this sample is representative,-to compute the number of

. residents and transients without private transportation. Third, KLD concedes that it.

has made.no computations with respect to populations of special facilities or private citizens with medical needs located within the EPZ. KLD #7, p. 1. . Fourth, KLD fails to include within its population estimates the substantia 1' number of individuals traveling through the EPZ, including the beach areas, at the time notification of an emergency may be given. KLD #2, p. 9. Accordingly, on its face, the KLD ETE lacks adequate data to compute the number of individuals or vehicles to be evacuated from the EPZ during an emergency. .W ithout such reasonably adequate data, therefore, KLD's computations regarding time estimates to complete evacuation must seriously be called into questiion.

s

o

? \i

, 3

'Forfreasons set forth above, the:KLD-ETE fails to provide reasonable assurance'that adequate, protective measures can or will be

. implemented in'the event'of radiological-u emergency.-

. Revised: Contention III of the Town of'Hampton to Evacuation-Time Estimate Report by KLD Associates,.Inc., May 23, 1986, at.

pp. 4-10, admitted per Board Memorandum and-Order of July.16,

'1987'at 7.

.(A)ipopulation' Estimates.

KLD lacks adequate data to compute'the permanent and transient population for the TownLof-Hampton'since KLD computes beach population capacities by examining only the.

beach above the high tide line, Vol. 6,

p. 2-12; counts parking spaces rather than motor. vehicles, including vehicles in transit, Vol. 6, p. 2-1; counts beach blankets rather than people, Vol. 6, p. 2-12; utilizesLa vehicle occupancy rate of 2.4 based upon two " field' surveys" performed on.

Weekends of frequent rain and poor beach weather, Vol. 6,.p. 1-10; and counts beach populations.using a limited number of photographs, of unspecified dateLor time, although KLD concedes'the beach populations vary widely. depending on weather,Ltime'of day,-andl day to day.- Vol. 6, p. 2-10. KLD thereby unreasonably reduces the actual population for the' Town of Hampton and distorts this " critical" factor in computing ETE. Vol. 6, p. 2-1. l (B) Weather Conditions.

While recognizing that weather represents a." major factor" affecting ETE, Volume 6, p.

3-1.and 2, KLD concedes that it has " limited empirical data on the effect of adverse weather conditions to reduce ETE." Vol. 6,

p. 3-1. KLD proceeds to arbitrarily reduce the ETE for rain and snow conditions for the Seabrook EpZ by 20 and 25 percent respectively, vol. 6, p. 3-11, although KLD lacks any site specific data on the extent of delay caused by these road hazards. KLD "

acknowledges the " issue of ocean fog," yet i

l 11__ ___ __ _ _ z_ _. ._______:__ _

fails to provide any data on the impact of i fog on ETE, Vol., 6, p. 3-11, and fails to respond to RAC concerns regarding wind .:

changes, which.may require contingencies for i the redirection of evacuation vehicles, New "

Hampshire Response Actions to RAC Review, August, 1986,Section VI, op. 7, 12, (hereinafter RAC Review, August, 1986) with additional delay to traffic and an increase in ETE. KLD does not even estimate the -

effect of ice storms on ETE.

i (C) Road Capacities, j When computing the " major factor" of road i network capacity, and its impact on the time I required to effectuate an evacuation, Vol. 6,

p. 3-1, KLD makes numerous and unsupported assumptions including:

1. All roads will remain passable during evacuation. Vol. 6, p . 10-70. This assumption ignores the obvious and anticipated vehicle breakdowns, gas shortages, overheating of vehicles, roadways becoming impassable from snow or ice storms, gridlock between evacuating private vehicles, commuters, and emergency vehicles attempting to enter the EPZ, and snowplow operators who either refuse to plow during radiological emergency or are unable to reach their designated routes due to evacuation traffic congestion.

2. KLD assumes that the " recommended traffic control tactics are in effect." l Vol. 6, p. 10-70. This assumption is i insupportable in view of the avowed position l of the Town of Hampton, and other towns l within the EPZ, not to implement the NHRERP !

if called upon to do so. The assumption is ;

further insupportable following RAC Exercise !

Assessment and Review of the NHRERP which

" cast (s) doubt" on the State's ability to provide adequate evacuation transportation.

RAC Review, August, 1986,Section VI, p. 9. j Finally, the State has failed to demonstrate '

an ability to provide sufficient law enforcement and traffic control personnel, FEMA, Final Exercise Assessment, 6/2/86 at l

l l

i.

I- ' ' '

p. 46, tofcompensate for non-participating towns. .Accordingly, KLD's1 reliance lupon thef State to supply adequate. equipment and Jpersonnel forLtraffic control management is

-without reasonable foundation.

3.- KLD assumes ~that 3,000 "through" vehicles will beitraveling through the EPZ at

'the time'of notification of an emergency ~.

Vol. 6, p. 10-3. No. support is provided for this assumption, which is rendered absurd by KLD's own? calculation of'" peak' hourly flow" on'I-95 of 6,912 vehicles. Vol. 6, p. 3-11.

Since I-95 represents only one. road-within the 200 square mile EPZ, the 3,000 "through" vehicle estimate represents a gross distortion of roadway demand.

4.. KLD-concedes that its " estimates of available capacity may overstate the actual accessible capacity." Vol. 6, p. 10-70. KLD thereby admits that its highway capacities relied upon'to compute ETE, and which represent.a " major factor" to calculate the time required for' evacuation, Vol. 6, p. 3-li would' generate.an unreasonably low ETE, and would not reflect actual conditions.

5. KLD unreasonably assumes-that 25 <

percentoof-the EPZ population will spontaneously. evacuate, Vol. 6,.p. 10-3, and estimates Hampton employees who work at the beach, both during.the week ~and on weekends, Vol. 6, p. 5-6, apparently by simple guess work. KLD thereby lacks adequate data to

~

compute road demand for Hampton. employees during e"acuation or to compute ETE when ~

partial evacuation of the EPZ is ordered.

6. KLD fails to adequately account for the impact of disabled vehicles on reducing ETE. Given.the thousands of vehicles to be evacuated, numerous disabled vehicles must be anticipated. KLD's claim that such vehicles will simply be pushed.aside by evacuees, without impacting on ETE, is insupportable.

Vol. 6, p. 12-4.

1 I 1 i V

u.

(D) ETE preparation Time.

3 ,

Without statistical-support, KLD assumes y that 90 percent-of the EpZ population will be

' notified of an emergency;within 15 minutes,

, Vol. 6, p. 4-8, assumes that beachgoers will be able to. leave the beach and access their ,

cars within 30 minutes, Vol. 6, p. 4-12, /

although.KLD concedes it has "no empirical up data to support.this distribution," Vol. 6, J p., 4-11, fails to allow for " staging area preparation time" as recommended by the RAC ', s in computing ETE, RAC Review, August, 1986, '

Section VI,.p. 10,.and grossly underestimate;es

'the adverse impact on ETE of 95 percent of . <

workers returning home, within 30 minutes, to e -

prepare for evacuation following notice of radiological emergency. Vol. 6, p. 4-9.

~~-l Moven.ent dish:ouragery Light.; '~~

~~O Traffic guide y,'~~

~~Ch '",faffic cone 5 e s~~

.. i X . Traffic barricade 5 -

~~DESC7IPTION: 1. Discourage eastbound movement along Route 110 and northbound movement along I-95.~~

~~2. Facilitate movement of traffic from T.oute>110 onto' southbound I-95. -~~
~~3. Permit eastbound/ movement along Route 110 if Macy M reet y mat #0WCH/LollIPMEhT approach to I,495 is not congested.~~
('4 , traffic 9tidos 101 traffic cc es ,
12 ' traffic barricades I-19 te ,; .
.3 ,
]
(1
.. ' il
(; >
r
?!
l f I
.1 L_
f.
i
! $' t
[. .
l ATTACHMENT 4 l
~~i. ( ; --~~
,f
'l 3
I i'
I i
i.
1

~~._.m_ _ . - _ _ _ _ . . . _ _ _ _~~

k.
Traffic Control Post No. B-AM-06
[with modifications] ERPA: B TowtJ : Amesbury LOCATION: I-951&. Reute 110 Nooc 253, 255 k
[ I-95 N
-s, 'N
\
X x x, . fm J/ %.
To I-495-
.Macy Street Route 110 4o
\-\
ok" , g '
_ gj j b\ \
N A _
QA
- -oMo-@ ooo
_-- c From x y _ _- - . .
-< = S alisbt.
Lic,je jx -W f -- % , , ,
e 7 .
N'%
kCbMa Suyts$C A) havemed -
g: pJry
> Movement facilitated 'Very
>l Movement discouraged Light O Traffic guide O Traffic cone X Traffic barricade DESCRIPTION: 1. Discourage eastbound movement along Route 110 and northbound movement along I-95.
~~2. Facilitate movement of traffic from Route 110 onto southbound I-95.~~
~~3. Permit Mbound movement along Route 110 if Macy Street MAMPOWCH/COtIIPMENT approach to I-495 is not congested.~~
4 traffic quides I 10 traffic cones 12 traffic barricades I-19
- --. , p .-- m -- - , - - .r.-----, -
+-c.--- . - - - y...,--- . .---,
-- ---- . - - - - - - - . _,-,--,-._,--m-----------,----,-.----- ,.,
l', f U g
' . f; , 4 4
5 I
~~13.-~~
n , ' 9,
, i i
<q-
. l.
(
ATTACHMENT 5 r.
Traffic Control P'ont No. D-HA-02 ERPA: D i p .' TOWN: Hampton i' . LOCATION: Interchange of Route 51, Route 101C & I-95
~~NODE: 230,233,235,236,237,238 I~~
h I-95 l c
N
/ ,_ d' l
% %h A.? N
~~k 4 Route 51 W~~
[
o e Q'-:%,
'~
r.
~
OcCoo
~saam *nm N o b *o o,
N
~~.::c: . - - - - ,~~
O
% * * *' . . . , _ _ k g: ,,,... -
I-95 ',
- I
>- Movement facilitated , ,
> I Movement discouraged -
y
@ Traffic guide -
O Traffic cone -
Route 51
)( Traffic barricade DESCRIPTION: 1. Facilitate movements from Route 101C to Route 51 west.
2.
~~3. Facilitate traffic moving from Route 51 to "I-95 nort 'I Facilitate traffic westbound on Route 51.~~
~~4. Discourage traffic from exiting Route 51 onto f MANPOWER / EQUIPMENT Route 101C and from northbound I-95 onto Route 51.~~
5 traffic guides l 18 traffic barricades 40 traffic cones I-39 I
4 I
,.h tr'i.
6 p
I I
i)
ATTACHMENT 6 i
l
^
l l-C . . i
f' L
n Traffic Control Post No. F-EX-04 ERPA: F TOWN: Exeter LOCATION: Exeter-Hampton Expressway & Exeter Rd. Interchange NODE: 47, 211, 212 h
Exeter Rd. j (Route 101D) ,
N EX _ ..
m tog ,0 q
sx N ,e, 0 resswa /
M9ht b M1 X Truck N '%
h j5 -
- +
/
o' o
% H m ton pg, oute 101c Key,:
> Movement facilitated
>l Movement discouraged O ' Traffic guide O.' Traffic cone K Traffic barricade DESCRIPTION: 1. Discourage Westbound traffic from getting off the expressway.
~~2. Force eastbound traffic off the expressway onto Exeter Road. l~~
_ MANPOWER /EOUIPMENT 3 traffic guides 7 barricades 6 traffic cones I-74 u___---____ - - - _ _
t
- $ ,g,- _ ,
~~r f~~
~~ii I~~
l' l-l I
. ATTACHMENT'7 1..
1'
Traffic-' Control Post No. B-SA-06 ERPA: B TOWN: Salisbury LOCATION: Elm St. (Route 110), Bridge Rd., Pleasant St., Beach Rd. (Route 1A), Lafayette Rd. (Route 1) & School St.
NODE: 93 f Route 110 to I-95 O . N Mudnock O Road' O Route 1 N s N N-o0 0 0
School St. 9
~~O g9 o g O o "~~
O N O \
o\
Routo 1, o
O M: 9 Pleasan.t St. .
o
>- Movement facilitated Route 1A
>-1 Movement discouraged
@ Traffic guide \
O Traffic cone \
X Traffic barricade DESCRIPTION: Split the traffic moving west on Route 1A onto two lanes approaching the intersection. The two . lanes will move continuously onto Route 110. All traffic moving south on Route 1 will be diverted westbound to Route 110. Northbound traffic on Route 1 will be routed to Route 110 via School Street.
MANPOWER / EQUIPMENT 5 traffic guides 32 traffic cones 3 traffic barricades I-28
\}h, l-ATTACHMENT 8
I AdbrPre-File: Attachment 1 9/3/87 Classification 1: Persons per . Vehicle i 1
Objective .
The objective was to determine the effect on ETE's of a vehicle occupancy of 2.3 persons for the evacuating permanent residents of the Seabrook EPZ. The ETE's in Volume 6 are based on a permanent resident vehicle occupancy value of 2.6. .
Approach The number of evacuating trips by the residents of each town was changed by a factor of 1.13 (= 2.6 + 2.3). An assignment and simulation were done for an evacuation of region 1 (the entire EPZ) under scenario 1 (summer, weekend, mid-day, good weather) using the revised number of permanent resident evacuation trips.-
Results The following table ecmpares the Volume 6 ETE's for the entire EPZ and the 2-mile ring with the ETE's resulting from a vehicle occupancy of 2.3.
ETE EPZ ETE 2-mile Volume 6 6:15 5:55 2.3 ppv 6:40 6:20 Classification 2: New Ilampshire Population Growth Objective The objective was to determine the effect on ETE's of projected population growth in the New Ilampshire towns that are within the Seabrook EPZ.
Approach The number of permanent resident evacuation trips for each New Hampshire town was increased proportionate to the projected town populations for 1990,1995,2000, and 2010.
Assignments and simulations were done for an evacuation of region 1 under scenario 1 using the revised number of permanent resident evacuation trips.
AdlerPre-File: Attachment ' 2 9/3/87
~~Results The following table documents the ETE's for the entire EPZ for the evacuation of region 1 under scenario 1, using the projected 1990,1995,2000 and 2010 New Hampshire town~~
_ populations.
Yr. ETE EPZ 1990 6:45 1995 6:55 2000 7:45 2010 9:40 Classification 3: Signal Timing Changes V Objective The objective was to determine the effect on ETE's ofintersection signal timings that are different from those used to determine the ETE's documented in Volume 6.
Approach The timings were equalized for the phases of each signal cycle spec! fled in the IDYNEV inputs for the evacuation of region 1 under scenario 1. An assignment and simulation were done for an evacuation of region 1 under scenario 1 using the revised signal timings.
Results The resulting ETE for the entire EPZ was 9 hours1.041667e-4 days 0.0025 hours 1.488095e-5 weeks 3.4245e-6 months and 5 minutes.
Classification 4: Delayed Implementation of Traffic Control Procedures Objective The objective was to demonstrate the effect on ETE's of delayed implementation of traffic L control procedures.
AdlerPre-File: Auachment 3 9/3/87 Approach The traffic control procedures documented in Appendix I of Volume 6 for the intersections ofI-95 and 110 in Amesbury and I-95 and 51 in Hampton were not fully implemented until I hour and 45 minutes after the order to evacuate in a simulation of an evacuation of region 1 under scenario 1.
I Results l The resulting ETE for the entire EPZ was 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months and 40 minutes.
Classification 5: Interstate On Ramp Capacity Reduction Objective The objective was to demonstrate the effect on ETE's of a capacity of 660 vehicles per hour for the I-95 on-ramp corresponding to nodes 45 to 255 of the evacuation network documentedin Figure 1-3 of Volume 6.
Approach ;
The discharge headway for link 45 to 255 of the evacuation network was set to correspond to a link capacity of 660 vehicles per hour. An assignment and simulation wm done for an evacuation of region 1 under scenario 1 using the revised on-ramp capacity.
Results The resulting ETE for the entire EPZ was 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months and 50 minutes.
Classification 6: Circuitous Routing Objective The objective was to demonstrate the effect on ETE's of: (1) New Hampshire beach evacuees taking direct routes to Manchester and Dover rather than taking the circuitous routes of Route 4 in New Hampshire and Route 236 in Maine; and (2) certain Portsmouth and Rye residents evacuating towards Dover and Rochester via the Sptulding Turnpike instead of taking the circuitous route of Route 236 in Maine.
l
AdlerPre-File: Attachment 4 9/3/87 Approach The origin / destination pairings for the New Hampshire beach load nodes and for certain .
I Porstmouth and Rye permanent resident load-nodes were changed in the IDYNEV '
assignment inputs for the evacuation of region 1 under scenario 1 in order to minimize circuitous routing. Assignments and simulations were done for: (1) the reduction of l circuitous routing for New Hampshire beach evacuees; and (2) the reduction of circuitous ,
routing for New Hampshire beach evacuees and some Portsmouth and Rye permanent residents evacuees.
Results The following table documents the ETE's for the entire EP'Z for the evacuation of region 1 under scenario 1 under the conditions described above.
ETE EPZ Direct Routing: NH Beaches 6:40 Direct Routing: NH Beaches &
some Ports & Rye Residents 6:55 Classification 7: Alternative Beach Host Locations Objective The objective was to demonstrate the effect on ETE's of a failure of some beach evacuees to go to their assigned host location.
Approach Two categories of analyses were done in which: (1) some of the Hampton beach population south of Route 101E evacuates to the south via Route 1 A; and (2) a large portion of the Hampton beach population south of Route 101E accesses I-95 at the Route 51/101C/I-95 intersection in Hampton. In the first category, analyses were done in which 1/3 of the beach population goes south via Route 1 A and in which 2/3 of the beach population goes south via Route 1 A. In the second category, an analysis was done in which 2/3 of the beach population accesses I-95 at the Hampton intemhange. Assignments and simulations were done for an evacuation of region 1 under scenario 1 for these conditions.
- . _ _ _ _ _-____________ _ _ i
AdlerPre-File: Attachment 5 9/3/87
. Results The following table documents the ETE's for the entire EPZ for the evacuation of region 1 under scenario 1 under the conditions described above.
ETE EPZ IS south via Rte. l A 7:10 2S south via Rte.1 A 9:10 2S. access I-95 at Hampton 7:10 Classification- 8: . Increased Beach Populations Objective The objective was to demonstrate the effect on ETE's of increased populations at the beach areas in the Seabrook EPZ.
Approach The number of evacuation trips generated by the beach areas was increased to conform to the capacity levels documented in the testimony of Doctors High, Befort, and Adler. An assignment and simulation were done for an evacuation of region 1 under scenario 1.
Results The following table documents the ETE's for the entire EPZ and for the Plum Island, Salisbury, Seabrook, and Hampton beach areas. Beach area ETE's are the times necessary to evacuate the areas immediately adjacent to the Atlantic Ocean and do not include the 7
additional time needed by beach evacuees to clear the EPZ. Volume 6 estimates are included f for comparison.
ETE (Volume 6)
EPZ 9:25 (6:15)
Plum Island 4:25 (3:05)
Salisbury 6:20 (4:45)
Seabrook 7:10 (5:40)
Hampton 8:45 (5.40)
AdlerPre-File: Attachment 6 9/3/87 Classification 9: Combined Effects Objective The objective was to demonstrate the eff~ t on ETE's of various combinations of the
]
conditions previously described.
i Approach The following three analyses were performed: )
(1) an evacuation of region 1 under scenario 4 (summer, weekday, mid-day, rain);
(2) an evacuation of region 1 under scenario 1; and (3) an evacuation of region 10 (beach amas) under scenario 1.
Analysis 1 included increased beach populations, delayed implementation of traffic control procedures at locations in Hampton and Amesbury, reduced capacity of an I-95 on-ramp in Amesbury, access of I-95 in Hampton by some beach evacuees from Hampton, and commuter flow at intersections in Amesbury, Hampton, and Portsmouth.
Analysis 2 included increased beach populations, reduced capacity of an I-95 on-ramp in Amesbury, access of I-95 in Hampton by some beach evacuees from Hampton, and commuter flow at locations in Amesbury and Hampton.
Analysis 3 included increased beach populations, reduced capacity of an I-95 on-ramp in Amesbury, access of I-95 in Hampton by some beach evacuees from Harnpton, commuter flow at locations in Amesbury and Hampton, and 50% spontaneous evacuation by workers and residents outside of region 10. )
Results The following tables document the ETE's for the three analyses. Again, the Plum Island, Salisbury, Seabrook, and Hampton beach area ETE's are the times necessary to evacuate the areas immediately adjacent to the Atlantic Ocean and do not include the additional time needed by beach evacuees to clear the EPZ. Volume 6 estimates are included for comparison.
i AdlerPrefale: Attachment 7 9/3/87 Analysis 1: Recion 1 Scenario 4 ETE (Volume 6)
EPZ 12:15 9:45 Analysis 2: Recion 1 Scenario 1 l
\
ETE (Volume 6)
EPZ 12:15 (6:15)
Plum Island 5:00 (3:05)
Salisbury 11:15 -(4:45)
Seabrook 7:30 (5:40)
Hampton 9:25 (5:40)
Analysis 3: Recion 10 Scenario 1 ETE (Volume 6)
Plum Island 4:15 (2:10)
Salisbury 10:10 (3:45)
Seabrook 5:15 (2:45)
Hampton 8:15 (3:40)
Classification 10: Commuter Flow Objective The objective was to demonstrate the effect on ETE's of commuter flows that will conflict with evacuating traffic and compete with evacuating traffic for roadway capacity.
Approach A limited number of commuter trips were loaded onto network links or routed through network intersections to simulate the effects of commuter traffic. One location modeled in this fashion was the I-95 on-ramp in Hampton. Three analyses were performed under the general conditions described in Analysis 2 of Classification 10 (an evacuation of region 10 under scenario 1). The analyses included the loading of 0,100, and 200 commuter trips onto the on-ramp over the period of time that commuter trips are expected to occur.
--- _ _ _ _ _ _ _ _ _ _ _ _ _ _ .___-n_______-_ _ _ _ _ - _ _ _ _ _ _ _ . _ _ _ _ . _ - _ . _ - - - - - - _ _ _ _ _ _ _ _ _ - _ _ - . _ _ _ _ . . . , _ _ ._ __ _ _ _ _ _ _ _ - - _ _ _ . - _ _ _ _ _ . _ - _ .
o c Adler Pre-File: Attachment 8 '9/3/87
~ Results .
The following table ' document the ETE's for the Hampton beach area and for the northern pan of the EPZ.
EIE EPZ north. Hampton beaches
. Analysis 1: O commuter trips 9:45 9:15 Analysis 2: 100 commuter trips. 9:50 '9:25
. Analysis 3: 200 commuter trips 9:55
.' 9:30 1
.}
o i
l
)
J
_ _ =_- :___
i
'r 7 -
l l
l ATTACHMENT 9 r
.l;
e L Access. Control Post No . NW- 1 TOWNi Newington PRIORITY: 3 LOCATION: At intersection of Woodbury Avenue and Gosling Road-
. (Portsmouth City boundary)
Woodbury Avenue , oooo L
/
- \
M
+ -
/ /////r_.- - -Va_
h 0
_J_ / / / / / /
X~ _ _X __ - g O_ __ _ _ _ _
>l O O
, r Exxon y Mobil Woodbury Avenue Station Station Gosling Road N
=
D Town of Newington City of Portsmouth
=
= Movement facilitated
=; Movement discouraged O Traffic guide O Traffic cone
,X Traffic barricade l
DESCRIPTION: 1. Restrict traffic movement southbound along Woodbury Avenue toward Portsmouth. j
~~2. Facilitate merging of traffic to the right on j southbound Woodbury Avenue, for right turn onto ;~~
Gosling Road.
MANPOWER / EQUIPMENT 3. Give priority to northbound traffic on 2 traffic guides #Y ^ * **
2 traffic barricades 10 traffic cones L ,t8
W .' . -
p i
h ATTACHMENT 10 I
h
= Table 1:
Example 'of Roadway Ch'aracteristics.
f1 c, Nucher 1 Segraant Type 2 Capaci ty3 Comments" of! Lanes L.
1 I
-NOTES: Total number of through lanes in both directions. If roadway cross section is not uniform, use section with least number of -
lanes 27 = Freeways and Expressways
'O = Urban Streets P,= Rural Highways 3
If known
" Indicate any special conditions that may affect roadway capacity.
i 4-15 I
I
7,----
. 1 l
o 4 1, 1 4
l .'
i I
')
- ATTACHMENT 11
e0 sh i a c .
it t r . m c r 4 a t t t e y a a y a a e aS f c t ht le t f v
,s t l o r i . i o E s s a uk o c) c c e a e po .f a r t a .
o pa p ro e el m t . pt p t .
y r b e ck pr a e c y r e s ac pc r r o a ah b e c u e m l l t h a c e o
ye h xm h i ph i 5 w . t e y5 t t m c af aS 7 l r t a s l 2 a a E a v o e k ri c e 5 e r e e e a b t t e a t p w e a t w w f a a e e n o u e a
.b l l h w ye0 ) i c y ) )
yhl a t t n - 0 n r a s n w t t ih i n od f e1 i a v f e i o ii f we ii oh a nE ~ oey a n c
a w h d t m a s
. t t a
r e c
r s s t si . ( .
o ( (
0 p s t i r c l f r t n S . t l 1 .
a n s w c wi u c c a uo p
u c
ce s ppe t
n rw
) c p p m t
n t
n
,. o r , c
- o . c o ey m e oo E e e 1 s t t u s oh pt o 0 f n 0 m m t a o e t c 5 n o e s 5 e e s n nTi ni t p n il l s( . l l o e ni t i w s i t a p c a t s c c m o i a i 0 a a mm n e e r at n n i m ik .l rt r0 r eE i n i i r o t e si n u1 nr mh n e a C a e n c ne o n o r a t oi r r n l ww a e r t t e ih .
o s a i s o o u - of d r a d t c s f e t n f f e pd t d u d d ai t ew a a c oi t u c n e u l h n t h l r t t S p mgh n g s
. S n o
a r a
S uw p ei b u t t n
u t p b u
b u
c
~
n af ii a a o s s s e . o a o e on ne .
e y e s r e e
. pe n ,
n ms pe ,
r i r r i a e o e y r . e e i a . a e a vt a ey v t t r v i l a t a y v v t t m v oi o si t o t c w s a o o a hc c pr e oh b c h o c a ap a cl b
a i l b i n e i h d r c -
b b u ri a a d i v a a c a d a p u c e n i u ad a e0 n n e s a em s oa r s o s r o ei s B 7 i ab A c B E A Tf a A Ci d T b m A A v
_ E -
f r e
_ o h t d n d n d n w d n
_ n a e
o i a
o i a
o i o o i o o o o a n o a W R G R G R R
_ i G S' G S t
u
_ b y y y y y y y g y g y i a a a a a a a n a n a r e d d d d d d d i d i d i t
m - - - - - - - n n
_ i d d d d d d d e el s T i i i i i i i v '; vl i M M M MM MM E A E A E
_ D
_ k k k k k k k k
. d d e e e e e ed ed
_ 1 n n e e e e e e n e n y
a k e
k e w w w w w we
- k we
- k w
0 e e D
_ 1 d d d d d d e d e d e e i i. i i i i e i e i e
_ e W W M MM MM MW MW M l
b a n n n n n T o o o o o
_ n s s s s s
_ o a a a a a
_ s r r r r e e e e e e e a e e S
_ S S S S S e m m m m - - - - -
S m m m m f f f f f f u u u u f f f f f f S S S S O O O O O O o
i r
a 1 2 3 4 5 6 7 8 9 0 n 1 e
c S
H?"
,l ,
,1 i
.1 i
3
'i ATTACHMENT 12 l
l 4
l 1
1 l
l l
1
i 1
'Y -
~~j 1~~
g V 1 x 1 Table 10-2. Identification of the Seabrook Station i
) Emergency Planning Areas (ERPA) l
-)
Recion Scatial Extent CRPA Designation 1 To EPZ bdry. A - G. Entire EPZ I
2 To EPZ bdry. 'A,lD, G Entire North Region
-3 To EPZ bdry. A, C, F Entire' West Region :
4' To EPZ.bdry. A, B, E Entire South Region-5 To Five Miles A,.B, C, D Entire Five-Mile i Region ,
6 A, D Inner North Region j ( To'Five Miles 7 To Five Miles A, C Inner West Region 8 To Five Miles A, B Inner South Region b
9 To Two Miles 2 AO Entire Two-Mile f ,
O Region ,.-
,% /'
10 Beach Areas Portions of- Beach Arkas ta,B,D,E,G !
?
1 NOTES: i. N11beac3areasarealwayscompletelyevacuatedin
, Scenario.1 1-4, including those outside the Region ordered to evacuate. It is assumed that beaches are s elosed at theiAlert sta'ge of the Emergency and that evacuation of the beach areas begins 20 minutes
.;,before the ordor to evacuate a specified region.
~~2. It is assumed that 25 percent of the population within the EPZ, but outside the Regien ordered to evacuate, will spontaneously evacunte, contrary to ,~~
)
instructions. It is also assumed that s i an additional 3,000 "through" vehicles, not otherwise !
countad, are on the highway when the Alert is announced.
~~3. The outer boundaries of Regions 5-9 are generally "~~
town boundaries which extend,7se.newhat, beyond the indicated distances from Seabrook Station. Thus, for l each of :hese . regions, the indicated radius is an 10eroximLaign and should not be interpreted
-literall!.
L 10-3 i
g
, q (4-
/
i L 1.
UNITED STATES OF AMERICA NUCLEAR REOur ATORY COMMISSION i
Before Administrative Judges:
Helen F. Hoyt, Chairperson Gustave A. Linenberger, Jr.
S Dr. Jerry Harbour
)
)
In'the. Matter of )
r )
PUBLIC SERVICE COMPANY OF NEW ) Docket Nos.
THAMPSHIRE, ET AL. ) 50-443-444-OL (Seabrook Station,. Units 1 and 2) ) (Off-site EP)
) September 14, 1987
)
TESTIMONY OF THOMAS J. ADLER ON BEHALF OF THE ATTORNEY GENERAL OF THE COMMONWEALTH OF MASSACHUSETTS ON REVISED SAPL CONTENTIONS 31 AND 37 AND TOWN OF HAMPTON REVISED CONTENTION III TO REVISION 2 (THE ETE CONTENTIONS)
AND ATTACHMENTS Department of the Attorney General Commonwealth of Massachusetts One Ashburton Place Boston, MA 02108-1698 (617) 727-2265 l
u_-A-.-
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION Before Administrative Judges:
Helen F. Hoyt, Chairperson Gustave A. Linenberger, Jr.
Dr. Jerry Harbour
)
)
In the Matter of )
)
PUBLIC SERVICE COMPANY OF NEW ) Docket Nos.
HAMPSHIRE, ET AL. ) 50-443-444-OL (Seabrook Station, Units 1 and 2) ) (Off-site EP)
) September 14, 1987
)
TESTIMONY OF THOMAS J. ADLER ON BEHALF OF THE ATTORNEY GENERAL OF THE COMMONWEALTH OF MASSACHUSETTS ON REVISED SAPL CONTENTIONS 31 AND 37 AND TOWN OF HAMPTON REVISED CONTENTION III TO REVISION 2 (THE ETE CONTENTIONS)
AND i
ATTACHMENTS
)
Department of the Attorney General d Commonwealth of Massachusetts !
One Ashburton Place Boston, MA 02108-1698 (617) 727-2265 i I
~~4 y p~~
,t. i e pf a.
/ i f
CN,ITED STATES OF AMERIDN ,j ' .
e ;
Bef$r.h Administrative Judges: )
' Helen g. Heyt, Chairperson GurtaW A. Linenberger,'Jr.
l J -dt. Jerry Harbour i
!}'
{' t t ) -
' ' .,I
-; ) 7.! .
In.-the Mattetnof )
)
~~q R PUBLIC. SERV (CE COMPANY OF NEW ) Docket Nos.~~
HAMPSHIRE, E"!1 AL. ) 50-J43-444-OL (Saabrook Station,, Units 1 and 2) ) UOffLoite EP):
t 4
)
s Spptembar 11c,1987 <
.) 7 s .
'l ,f f s/ TESTIMONYl OFaTHOMAS J.-ADLER ON BEHALF )
., c .OF THE AT QPNEY GENERAL OF THE' COMMONWEALTH ,I t
OF MASSAci10SETTS ON REVISED SAPL CONTENTIONfy 31 ANDl,,37 AND TOWN OF HAMPTON REVISE'S
~
n 3 -,1 , CONTENTION III TO REVISION 2 (THE ETE CONTENTIONS)*
-]> Tha Saabrook Ei nergency Planning Zona, or EPZ, includds communities out to ten mi.las away from Saabt;och Station y j +
extanding north / south f rom Portsmouth, M4w HattpshirS to Nawbury, Massachusetts and to tha west intoExsQr, Kingston and Brentwood, Naw Harnshire. However, Sha 'tew Hampshire and t J Massachusetts seacoast' area constitutes tina'most important part of the.Seabrook Emergancy Planning Zona. Iraportant becaqst of l 4
i "tha nurabar o6.paopla and cars in the araa throughout.tha summar,.
1 y y .
This tastimony, is i d< transcript 4 tha narrativa- of tha videotape attached.h9ta%as 1A%adms nt 1. The affidavit of Kant Batclay, thW individus1 9ho cigatad that vidaotapa, is Attachment 2.
t s
A 1
L 1 i
L i 1
M.
i months and also important bacause of the limitad roadway systam that serves tha area. Rt. lA, known also as Ocean Boulevard axtends. north and south out of Seabrook. To tha north it passes along the crowded beach area in Hampton than winds along tha coast'through North Hampton, Rya-and, finally, into downtown Portsmouth. In part becausa of tha axtansiva salt marshas to the immediata west of the baach araa, there are only 5 major roads connecting to the west ovar the ten-mila stratch between Saabrook and Portsmouth.
- Rt. 1 or Lafayette Road runs north / south paralleling Ocean Boulevard and forms a major I junction at the Portsmouth traffic circle. Interstate 95 passas north and south through tha arma 4 miles to tha wast of tha baaches, but it has only two accass points ovar the ten
~
milas of highway passing through Seabrook, Hampton, North Hampton and Rya.
l Tha New Hampshire Radiological Euargancy Responsa Plan l statas that an Alert at tha Saabrook Station will trigger a j baach closing massaga for all baachas within 10 milas of the plant. According to tha Plan, baach visitors will, on haaring this beach closing massaga, pack thair belongings, assemble thair party and walk to their car. Those at Hampton Baach will face barricadas across Rt. lA to tha south and so will ba diractad to drive north along Ocaan Boulevard aithar to Highland Avanua, than to Rt. 51, tha Exatar-Hampton Exprassway or furthar north to Rt. 101C, High Straat. Tha ona and one-half mila trip from Hampton Baach Stata Park to Rt. 51 will, according.to tha Plan, taka up to 5 and 1/2~ hours on a hot summer weakend. Those who choose to turn wast onto Rt. 51 Will be routad across Rt.1 to I-95 whara thay will be directad into the aastbound lana and. discouraged from accessing tha I-95 on-ramp. From thera thay will continua wast along Rt. 101 to a decontamination conter in Manchaster. Those who' continue north along Ocaan Boulavard to High Straet will traval west to tha-sama I-95 on-ramp and will be diracted onto I-95 northbound to
.US.4, tha Spaulding Turnpika. From thara they will continue to I-93 and I-293 to tha Manchestar decontamination cantar.
At tha'baach closing, visitors along Seabrook baach will be routad south along Ocean Boulevard to Rt.'286. If Rt.
286 is not congasted, thay will ba diractad wast onto 286, across Rt. I and over to I-95 southbound. According to the New Hampshira Plan, this trip will take also up to 5 and 1/2 hours. From there thay will procaad to I-495 to Rt. 97 to Rt.
28 to a decontamination cantar in Salam. If Rt. 286 is congastad, Saabrook baach visitors will be directed furthat south along Rt. 1/Rt. lA/Rt. 110/ Plaasant St./ Mudnock Rd.
intersection, occupying both aastbound and wastbound lanas of Rt. 110. At tha I-95 interchange, half will ba diractad onto tha southbound on-ramp and tha other half will ba diractad onto what is normally tha southbound off-ramp, than ovar a raised, grassy madian and finally onto I-95 south to a decontamination centar.
3- j
1 1
Thase axampla avacuation routes illustrate savaral important' features of this araa: a singla road winding through the most dansaly populatad beach areas, a limited number of east-west connections off this road and an evan mora limitad number of access points to the area's major artarial, Interstate 95.- This road natwork was designed to serve the n'eads of an active saacoast community with constant flows of tourists along tha coast north and south. It was not designad to serva a nuclear avacuation and a major issue is how well it would, in fact, serva such a purposa.
l 1
ATTACHMENT 1 i
0 UNITED STATES OF AMERICA NUCLEAR REGULATORY-COMMISSION.
Before Administrative Judgas:
Halen F. Hoyc, Chairperson Gustave A. Linanberger, Jr.
Dr. Jerry Harbour
)
)
In the Matter of )
)
PUBLIC SERVICE COMPANY OF NEW ) Dockat Nos.
HAMPSHIRE, ET AL.- ) 50-443-444-OL (Seabrook Station, Units 1 and 2) ) -(Off-site EP)
) Sapt9mber 14, 1987
)
AFFIDAVIT OF KENT BARCLAY I, Kant Barclay, do make oath and swaar:
1). My nana is Kant Barclay. I am tha Production l
l Managar at Cagney's Vidao in Middlaton, Massachusetts. I have l
been producing and craating videotapes for ovar 8 years, for l'
I the last 5 years in the litigation support fiald, 1
2) On Septamber 6 and 7, 1987, I took videotapa footage of various roadways and intersections betuaen Portsmouth, New Hampshira and Nawbury, Massachusetts. Tha video camara and equipment I usad to film thasa scenas was normal vidao aquipment, all in good operating order. Tha aquipment I used is generally accurate in racording images.
1 1
1
3)' The imagas that appaar on the videotape submitted as an attachment to the narrativa tastimony of Thomas J. Adler, are videotapa images of tha scenes.that I witnessad during my filming on Septembar.6 and 7, 1987 A).1 of these images fairly and accurately represent tha actual conditions as I viawed them. I filmad approximately 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months of videotape which I adited to create tha videotape submitted here. I recorded Thomas J. Adler's voice and matchad the video images of various roadways and intersections I had filmed to his narrative.
SIGNED UNDER THE PAINS AND PENALTIES OF PERJURY THIS 10TH DAY OF SEPTEMBER, 1987..
\ \
'a' l Kent Barclay

ML20238F214

Text

SUMMARY

Navigation menu

Search