ML19331A769
| ML19331A769 | |
| Person / Time | |
|---|---|
| Site: | Midland |
| Issue date: | 05/18/1977 |
| From: | Feld S, Gundersen W FEDERAL POWER COMMISSION, Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19331A768 | List: |
| References | |
| NUDOCS 8007230779 | |
| Download: ML19331A769 (40) | |
Text
i-m
'(
ATTACHMENT A UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of
)
)
CONSUMERS POWER COMPANY
)
Docket Nos. 50-329
)
50-330 (Midland Plant, Units 1 and 2)
)
REPORT PREPARED BY S. FELD AND W. GUNDERSEN FOR THE ATOMIC SAFETY AND LICENSING BOARD ON ALLEGED DISCREPANCIES BETWEEN CONSUMERS POWER'S RATE FILING OF JANUARY 31, 1977 AND TESTDONY PRESENTED TO THE ATOMIC SAFETY AND LICENSING BOARD IN THE MIDLAND PROCEEDING
~
Introduction The Staff, as directed by the Atomic Safety and Licensing Board, has reviewed the rate filing of Consumers Power Company (Licensee) before the Michigan Public Service Commis: on of January 31, 1977. This rate filing is essen-tially a detailed justification for rate revision on the Consumers Power's system.
Ccunsel for All Intervenors Except Ocw (Intervenors) has charged that significant contradictions exist between testimony provided to the Atomic Safety and Licensing Board by the Licensee and representations made before the Micnigan Puolic Service Comissicn in the rate filing.E Clearly, similar information and analyses do appear in the information filed in both proceedings and certain differences in treatment can be observed. However, the Staff review concludes that:
M The allegations were made at the March 2a,1977 hearing session in this proceeding at Tr. pages 5038 through 5045.
77 7 8007230 1
9 i
~2~
,-~
1.
The discrepancies have a negligible effect on the conclusions-reached by the Licensee in this NRC proceeding and in no way would alter those conclusions; 2.
The discrepancies do not produce any systematic bias, i.e.,
where differences were observed, the different treatment in the NRC proceeding was not always beneficial to the Licensee's case relative to the treatment used in the rate filing; 3.
The discrepancies, with the exception of a few minor errors,
~
are fully explainable in that they either reflect (a) a basic philosophical difference in the intent of the information pre-sented, or (b) differences in the cut-off date used to. prepare the analyses which precluded the input of more recent information and management decisions.
I.
Staff Review of the January 31, 1977 Rate Filing The Staff has identified those areas in the rate filing that contain material relevant to testimony presented by the Licensee in the NRC proceeding and has reviewed them for consistency.
This review included meetings and discussions with the Licensee in order to obtain additional information.2_/
In all, five areas for review have been identified:
2_/ A meeting was held between the Licensee and the NRC Staff in Chicago on April 5, 1977.
The Licensee has provided the Staff with information it requested in letters of April 13, 1977 and May 10,1977 (distributed at hearing on that date) frcm the Licensee to the Board with ccpies to all parties in this proceeding.
s
. 1.
Need for Power - Forecasted energy and pcwer loads, and capacity planning, specifically treatment of sales to co-ops and municipalities, and plans to temporarily derate Palisades; 2.
Alternative Energy Sources - Coal fired generating plants, specifically projected coal prices; 3.
System Reliability and Reserve Margin Analysis - specifically random outage rates; 4.
Cost of Replacement Power (delay costs) - specifically planned capacity, fossil fuel costs, outage rates, scheduled maintenance, heat rate, sales to co-ops and municipalities, nuclear fuel costs, and use of different models.
5; Treatment of Decommissicning Costs.
1.
NEED FOR POWER - Forecasted Energy and Pcwer Loads and Capacity Planning The material in the rate filing pertinent to these issues appear in the direct testimony of Mr. E. H. Kaiser, !1r. J. fi. Brager, and Mr. P. L. Bickel.
A.
FORECASTED EMERGY AND POWER LOADS In Mr. J. M. Brager's testimony there is a discussion of crojected load factors, energy sales, energy requirements, efficiency factors,
4_
and peak load demands througn 1986.
Tnese values, as well as his discussion of them, are ccmoletely censistent witn the Licensee's presentaticn at the NRC hearing.
In Mr. P. L. 31ckel's testimcny there is a discussicn of forecasting methodology and projected gr wth through 1985 by major customer class. Tnis presentation is totally censistent with the March 15, 1977 Amencment to the Envircnmental Report and the direct testi=cny of P. L. Bickel at the NRC hearing.
In Mr. E. H. Kaiser's testimony, peak load demand forecasts are used in his Exhibit F2 to develop projected reserve margins.
Tne peak load forecast is identical to tnat used by Mr. Brager and in the NRC proceeding.
3.
CAPACIT( PLANNING With respect to ca acity planning, Mr. Kaiser's Schedule F1 a:: earing in the rate applicaticn reports major capacity additiens, changes, and retirements through 1986.
The plar.ned sale of 60 W of Cam bell 3,17 31 of Midland 2, and 98 W of Midland 1 are identical as between.he rate filinC and CP's capacity planning as described at tr,e NRC hearing.
Mr. Kaiser's Schedule F1 reports tha retirement of Morrow 1 and 2 in 1983.
This results in a reductin in capacity of 66 MW.
This decision was reached after the Licensee filed its testimony in the fiRC proceeding-and is not incorporated in the Licensee's case before the NRC.
However, if it had been, it would have resulted in a greater need for Midland than was presented by the Licensee.
Mr. Kaiser's Schedule F2 presents CP's net capability through 1985 -
the following differences are noted between this exhibit and values used in the NRC proceeding:
1.
Rate filing indicates an additional 11 MW derating on Palisades from 1977 through 1980 due to cooling tower requirements.
2.
Rate filing indicates a 10 fM derating at Big Rock from 1977 through 1984 due to ECCS limitations.
l 3.
Rate filing indicates retirement of Morrow 1 and 2 commencing in 1983 of 66 MW.
1 l
U The testimony of tne Licensee and the flRC Staff were filed in this proceeding on ilovember 5, %76.
Final decisions with regard to the rate case app'..ation were not reached by the Licensee until mid to end December 1976.
Updated material was not presented in the ilRC proceeding as'its effects were judged by the Licensee to be insignificant.
The Staff concurs in this conclusion reached by the Licensee.
s
. The above three items are not considered by the Licensee in the flRC proceeding because they were decisions made after the testi-many in the NRC proceeding was filed.
(See footnote 3).
- However, if they had been incorporated into the NRC proceeding, each item would have enhanced the Licensee's need for the Midland units.
Two additional discrepancies exist because the rate filing is estirating winter net capability whereas the NRC proceeding is con-cerned with the summer net capability.
They are:
169 ?",1 summer derate; and a 159 tiW reduction in net capability in 1983 because that portion of the Luddington sale terminates in August of 1983.
Thus, the rate filing shows higher capabilities because the sumer derate is not applicable 90 winter estimates and whereas the additional Luddington capacity will bc available for the winter peak of 1983, it will not be availabic for that year's sumer peak.
The Staff agrees that the sumer peak is t.1e relevant peak to use in the NRC proceeding to determine reliability and that these i
differences are fully justified. A discussion of the summer derate and termination of Luddington sale appear in the Licensees Environmental Report, p.1.1-20, and Table 1.1-8, respectively.
In all other respects, including the proposed derating of the Palisades Unit due to steam generator tubing problems (71MW in 1978, 35 MW additional in 1979 and 1980, and a complete outage in 1981 and 1982), the rate filing and CP's position at the NRC hearing are identical with rescect to capacity planni'ng relevant to need for power.
i
s
, 2.
ALTERNATIVE ENERGY SOURCES - Projected Coal Prices Projected coal prices are important to the NRC case in two major respects:
coal prices impact on the cost of Midland vs. coal fired alternatives, and coal prices impact on the cost of delay as used in the production cost runs.
This discussion is limited solely to the treatment of coci prices in the rate filing as they relate to the alternatives analysis. A more detailed discussion on coal prices as they impact on the costs of delay will be presented in Section 4 of this report.
For the alternatives analysis, the cost of newly contracted coal is the relevant consideration and the testimony by Mr. R. Wilkenson in the HRC hearino orovides the basis for these projected values.
The discussion on coal prices in the rate filing appears in the direct testimony of Mr. J. M. Brager and Mr. J. E. Van Reenen.
Their cencern is to project the expected averace cost of coal for each coal fired unit on the CP system.
These data are necessary to estimate revenue requirements which are, of course, critical to a decision on rate relief.
They are not, however, comparable to the estimates developed by R. Wilkenson.
Clearly, the averace cost reflects a myriad of factors in addition to the cost of newly con-tracted coal.
For example, the values presented by Brager and Van Reene must censider the mix between existing and new contract coal, the point in time at which new centracts replace old enes,
~.
, shifts frcm high sulfur to low sulfur coal, and shifts in coal supplies frca one plant to another.
Each of these factors will cause the average price and resulting escalation rate to vary frcm these used by Wilkensen in his testi=cny.
Table 1 presents the forecasted coal prices by generating plant for the period 1977 through 1982 as depicted in the rate filing. a /
The effect of these facters can be seen by reviewing several of the data points.
For example, the price of delivered coal to Karn 1 & 2 is ferecasted to increase by almost 55% between 1977 and 1978.
If an existing centract were to be in force between these two years,.cne would expect escalation of 12% based en Wilkenson's testimony. Mcwever, in actuality these units will begin switching to low sulfur coal in 1977 with ccmolete ccnversien by 1980.
Thus, the resulting value of 51.81 which appears in the rate filing reflects that portien of Karn's coal supsly subject to the higher price of this new low sulfur coal.
Alternatively, the rate filing shcws less than a di escalaticn rate in coal deliveries to Campbell 1 & 2 in 1978.
This escalatien is well below Wilkensen's assumed 12% escalatien because icwer,criced ccal frcm Weadock 7 & 8 will be diverted to Cacobell 1 & 2 in 1978.
I/ estimeny of J.!1. Brager of Censurers Pcwer Co. befcre the !!ichican T
Public Service Cc mission, January 18, 1977, Exhibit JM3 9.
s
.g.
Table 1 - Coal Prices Report in Rate Filing 1977 1978 1979 1980 1981 1982 (COST IN CENTS PER MILLION B.T.U)
Weadock 7-3 100 145 162 178 206 227 Xarn 1-2 117 181 199 218 241 265 Whiting 1-3 132 150 166 220 242 268 Campbell 1-2 106 110 121 134 147 162 Cobb l-3 100 137 154 194 218 241 Cobb 4-5 100 137 154 194 218 241 Campbell 3 232 256 302
~
For these reascns, it is not possible to completely reconcile Wilkenson's estimates on new source coal with the coal data presented in the~ rate filing.
Such a reconcilation is not achievable as the purposes for which the coal cost data are being used are different in the two proceedings.
For the rate filing, average coal cost data for each CP unit is required.
These coal costs reflect the many factors discussed above.
For the NRC proceeding, the appropriate coal costs for an alternatives analysis is the cost of newly contracted coal.
So while the coal costs differ in the two proceedings, the application of each is correct.
m m
It should be further pointed out that there is an explicit statement in Van Reenen's testimony that in developing his estimates, he assumed real price increases of 6% annually for 1977 and 1978, and 4% annually for 1979-1982.
These real price increases represent the exact real p-ice increases used by Wilkinson in his testimony before the flRC.
Thus, while absolute coal costs in the two proceedings may differ, the treatment of escalation is consistent.
3 - SYSTEM RELIABILITY AMD RESERVE MARGI!! AflALYSIS Investigation revealed that the use of random outage ratesEI by Consumers
~
Power Company for the development of required reserve margins was essentially consistent in both the NRC hearings and the January 31, 1977 rate filing.
In -
both cases, the impact of historic outage rates and proje' ted outage c
rates was evaluated to reflect the sensitivity of randem outage rate variations as related to the r',iability of the system.
The only randem outage rate variation between data employed for the rate filing as compared to the 111dland case was the assigned historic random outage rate of the Sig Rock (71. 1) unit. The random outage rate for Sig Rock used for the' 4
flidland hearings was 29.53 as compared to 18.65 in the January 31, 1977 rate filing. This historic rate of 18.65 is the correct number and its
- )
application in the Midland proceeding testimony would favor the delay of 5/
andom outage rates include forced outages and short-term maintenance R
Short-term maintenance outages are, in errect, rorced outages outages.
that can be postponed beyond the next weekend if necessary and, therefore.
do not affect operating reserve requirements but must be factored into required installed reserve calculations.
s
. the Midland Units but wculd be negligable adjustment resulting in a decrease in reserve cargin requirements of Only 0.225.
Other miner variations in input parameters for the less of 1 cad probability studies were:
1.
Slight variat.cns in tne clanned raintenance schedule 2.
Retirement date of Merrew Units 1 and 2.
Maintenance schedules are revised frecuently in = cst ;cwer systems.
One reascn is the impact of unfereseen forced cutages that cccur and, while the unit is shut down during the forced cutage, certain reutine =aintenance functions will be perfer=ed that will effectively reduce er delay the scheduled cutage at the later date.
This prudent practice is widely used because of the cbvicus ecencaic benefits. Another reascn fer revised maintenance schedules is the revised delivery dates of reclacement parts needed during the scheduled overhaul. The maintenance schedules used in the !!RC hearings vary slightly from these used in the rate filing for the years 1951, 1982, and 1963.
These =iner variatiens wculd not affect the required reserve margins.
Morr w units 1 and 2 were assumed to be in service thrcugh 1935 in the Midland creceeding LCL? evaluations while, in the rate filing, these two units were retired in 1953.
Since the total ca:acility of tne r.<c units is only 55 MN, tne resultant affect is negligible. Mcwever, the retirement
of the Morrew Units would favor centinued ccnstructicn of the Midland units and, therefore, was a censervative application in determining the need for pcuer in the Midland preceeding.
Statements made by Mr. E. H. Kaiser l before the Michigan Public *.:rvice r'
Cccmission en January 31, 1977, in regard to the acplication of randem cutage rates as they affect recuired reserve cargin are censistent with the methcdology employed by Mr. Gcrden L. Heins _/ before the Licensing 7
Board at the Midland Hearings and also consistent with the informatien centained in the Envirencental Report Supplecent.8/
4 - COST OF RE?LACEM '
POWER (DELAY COSTS)
Althcugh the rate filing not ccncern itself with the costs cf delay, it does provide much informaticn that is pertinent to the Licensee's analysis of this subject as presented at the !!RC preceeding.
- Clearly, there are certain discrepancies between the two cases.
Mcwever, as this review will attempt to der.cnscrate, ali of the differences are fully explainable.
Furcheracre, even if the rate filing da:a had been used instead of the data actually used at the ?lRC hearing, the effect en delay ccsts would havr. been negligible.
In fact, the Licensee has adjusted for most of the incensistencies and the end result is cnly abcut a 25 reducticn in delay cet.3.
6/ ? age 5, Direc: Testiceny of E.H. Kaiser JJ Paragrach II, ::c. S and 9, Testi=cny of Gordon L. Heins.
8,/ Table 1.1-10, Figure 1.1-2, :aragra:n 1.1.3.
~
m A.
TREATMEtiT OF THE SALE OF PORTI0fiS OF MIDLAUD AfiD CAMPBELL 3 CAPACITY Both the rate filing and tha ilRC proceeding acknculedge identical sales and buy-backs of portions of Midland Units 1 and 2 and Cacobell 3 to municipalities and co-ops. This censistent treatment was already discussed in Section 1 of this report.
Mcwever, when Censumers Power calculated the costs of delaying Midland 1 and 2, it did so with respect to the full
~
capah'lity of these units, independent of these sales.
The Staff agrees, tnat in order to assess the imcact of delaying the cceration of Midland, total output of the plant cust be censidered, because the delay of Midland will affect the total output of these units and the effects will extend beyond the CP system.
Therefore, what appears to be a discrepancy is, in the Staff's opinion, a basic philosophical difference in the purpose of the information presented.
3.
RAi! DOM CUTAGE RATES Soth the rate filing, and the produr.tien cost runs used in-the NRC proceeding rely en projected random outage rates.
Of the 27 fossil units on the CP system, the cutage rates frem both sets of data are perfectly consistent with respect to 18 units. Mcwever, for the remaining 9 units, slight discrecancies are noted. A typical dif-ference can be cbserved by reviewing the outage rates recorted for Karn 2.
These values are recroduced below:
c
. NRC Proceedina Rate Case Production Cost Run 1981 1982 1983 1984 1981 1982 1983 1984 Karn 2
.0744
.0744.0744
.0744
.0676
.0744
.0744
.0744 The values reported in the rate case are constant over the 1981-84 period whereas in the production cost runs the 1981 value is lower by.0068.
Where differences are noted, the difference reflects the fact that for the oroducticr. cost runs the randem outage rates as developed by CP's operations division were used whereas for the rate filing a constant value was used.
In any event, the effect of differences of such small magnitude on delay costs is truely minimal.
In fact, even if the differences were more significant the effect on delay costs would still be of minor importance.
This is so because the delay costs are calculated as the difference between production costs incurred with Midland on line as scheduled vs. Midland delayed.
Since the same random outage rates would be used in each case, the effect of using a difrerent set of outage rates in calculating delay costs would essentially cancel out.
y C.
SCHEDtJt.ED MAI!!TEf!Af!CE Both the rate filing, and the production cost runs used in the NRC proceeding provide estimates of scheduled maintenance.
Of the 27 fossil units on the CP system, the scheduled maintenance from both sets of data are perfectly consistent with respect to 23 units.
For the remaining a units slight discrepancies are noted.
For three of these cases the total number of weeks of scheduled maintenance are the same over the 1981-84 period with the discrepancy only occurring in the assignment of scheduled maintenance to a specific year.
In the case of Weadock 8, 4 additional weeks of scheduled maintenance were planned under the production cost runs.
For clarification, the differences are reproduced below:
NRC Proceeding Rate Case Production Cost Run 1981 1982 1983 1984 1981 1982 1983 1984 Camp 1 4
4 4
8 8
4 4
4 Wead 8 4
4 4
4 4
8 4
a Whit 1 2
2 2
6 6
2 2
2 Whit 3 2
2 2
6 2
2 6
2
i
,m
)
. The explanation for these differences rests on the availability of updated information.
(See footnote 3).
The new schedules were available when the rate case was ~being prepared but were not available at the time the Midland case was prepared.
The Staff recognizes that scheduled maintenance is an item constantly subject to revision and up-date.
In any event, the differences are very minor and because the same scheduled maintenance would be used in the Midland on line and Midland delay cases, the effect of using different scheduled maintenance values in calculating delay costs would be negligible.
D.
HEAT RATES Both the rate filing, and the production cost runs used in the NRC proceeding provide estimates of projected heat rates.
In the rate case, a model using actual 1975 heat rates was used to develop the projected heat rates whereas in the NRC proceeding a different model which utilized actual heat rates over the 1971-74 period was used.
The results of the two approaches produce a number of inconsistencies.
In all, heat rates between the two models are identical for only 2 units, the heat rates are higher in the rate case for 4 units, and lower for all other units. Although the differences are many, the size of the differences are small.
In all but two cases, values vary by less than 55, with most of these being in the 1 to ?% range.
The largest differences tend to occur among those units for which the rate case resulted in higher values for the heat rates.
s
, An i. ccrtant reascn for the differences is that each codel assumes different operating levels for the units and the heat rate will vary as a function of unit use.
In addition, the codels ther.selves have.
certain limitations with resoect to clotting the heat rate curves.
The fact that heat rates accear to be higher in the NRC preceeding produces higher producticn ecsts than wculd have occurred if the data frc= the rate filing had been used. Mcwever, ence again because the data are used in both *he Midland as scheduled and Midland delay cases, and differences are the focus of the cost of delay analysis, the net effect is minimized.
E.
MNH CUTFUT FOR FALISADES A:iD SIG RCCK The rate filing and the NRC preceeding centain ferecasts of the MWh cutcut of Palisades and Big Rock in the 1981-84 time period.
The esticates in the creduction cost runs overstate these cutcuts relative to the rate case.
For Falisades, the difference is minor, en the crder of-less than 0.55.
For 3ig Rock the difference is ac?roxi=ately 35.
To the extent these cut;uts are overstated in the NRC preceeding, the costs cf delay are understated.
m With respect to Palisades the difference is due solely to the manner in which Palisades output is entered.
In the rate case, Palisades MWh cutput is a given, provided by the nuclear production people at CP.
In the production cost runs this value must be calculated las a part of the ccmouter run.
The prcgrammer cannot exactly duplicate the desired number because the computer rules him.
In the case of Big Rock, the same problem must be overccme.
- However, that adjustment does not fully explain the much larger discrecancy which exists here.
Referring back to Section 1 of this report it is noted that the rate filing has taken account of a 10 MW derating en Big Rock that is not reflected in the NRC proceeding. This discrecancy exists because the decision to derate occurred after the NRC. case was prepared. This ccupied with the minor inaccuracy produced by the ccmputer fully explains the apparent discrepancy.
F.
NUCLEAR FUEL COSTS - PALISADES, BIG ROCK, AND MIDLAND i - Palisades - The nuclear fuel costs for Palisades are approxinately 4% higher in 1981 and 7% lower in 1983 and 1984 than those reported in the rate filing.
The discrepancies are due to different M'dh cutputs assumed in the two cases and the fact that these values are continually updated and the rate filf,np which was produced at a later date reflected more recent data.
Mcwever, ' adjusting for these differences will not affect the escitate of cost of delay.
wr, Palisades will operate at its maximun level possible independent of whether Midland is delayed or not.
Consecuently, Palisades cannot contribute to the incremental cost of delay.
ii - Big Rock - The nuclear fuel costs for Big Rock are significantly lower in the NRC proceeding than those reported in the rate filing.
For example, in 1984 the estimated cost is less than one-third that reported in the rate filing.
The exolanation is that the cost of Big Rock generation has not been up dated in the oroduction cost runs. This is justifiable because this data is unnecessary input to the cost of delay.
The output of Big Rock is fixed for the as scheduled and delay cases and thus, when the difference is taken, the costs will wash out.
til - Midland - The Midland fuel costs for 1981-84, exoressed in mills /KWh are significantly higher in the production cost run than those reported in the rate filing.
This is so, because the production cost runs reflect the updated nuclear fuel estimates of February 4,1977 that was submitted at the NRC hearing whereas the rate filing is predicated on the earlier estimates of November 1976.
If the values reported in the rate filina were used in the production cost run, the cost of delay would increase.
Thus, relative to the rate filing, these values are conservative in their estii.ation of delay costs.
t G.
PROJECTED FOSSIL FUEL COSTS The cost of delaying Midland is a direct function of the estimated fossil fuel costs on the Consumers Pcwer system in the 1931-84 time period.
That is, the higher the fossil fuel cost estimates, the greater will be the delay costs.
However, because the sacre values are applied to both the as scheduled and delay cases, the effect of using a different set of values will be minimized.
Estimated fuel costs are reported in the Licensee's production cost runs used in the !!RC proceeding and in the licensee's rate filing before the MPSC.
Discrepancies between the two sets of data can be observed.
However, in all but five instances the differences are extremely ninor, on the order of 1" or less.
These differences are due to round off errors and a slight variation in how inflation and real price increases were handled:
e.g., rate case - assumino $1.00 per MBTU orice, a 12" price increase representing 65 inflation was calculated as --
$1.00 (1.06) (1.06) = $1.1236 whereas in the production cost run, with the very same assumptions it was calculated as --
i
$1.00 (1.12) = $1.12 In any event, because of these factors, the prcducticn cost runs produce slightly lower estimates than these accearing in the rate filing and thus, are conservative.
. More significant fuel cost differences are noted for Karn 1 and 2, Campbell 3, Weadock 1-6, Karn 3 and 4, and Gaylord Oil & Gas.
With respect to Karn 1 and 2, the production cost runs have about a 26%
higher coal cost for the 1981-84 period which tends to overstate delay costs. The lower value used in the rate case reflects a management decision, made after the NRC testimony was prepared.
(See footnote 3).
to contract for these units' coal socaly now at a lower cost rather than wait until a later date. The production costs runs on which the Heins delay costs are based are in error as they do not use the coal cost data for the Karn 1 & 2 Units presented in the rate filing.
With respect to Campbell 3, the production cost runs have about a 10%
higher coal estimate for the year 1984 which also overstates delay costs. The lower value, used in the rate case, represents the estimated cost of coal to Campbell 3 based on a more recent management decisicn (See footnote 3) to begin contracting for it now.
In the production cost runs, the estimate was based on an earlier management decision to contract for deliveries at a later date.
In addition, included within this estimate is the higher cost of coal to Campbell 4 which was scheduled to come on line in 1984.
This plant had been incefinitely delayed prior to the filing of Consumers testimony in the Midland proceeding. The production ecst runs on which the Heins delay costs are based are therefore in error as they do not use the coal cost data for the Campbell 3 unit presented in the rate filing.
,m
. With respect to Weadock 1-6, and Karn 3 & 4, the rate filing has about a 30% higher fuel cost estimate for the year 1982 which t ends to under-state the costs of delay.
In the production cost runs, the Morrow fuel costs were incorrectly inputed for these units whereas in the rate filing the correct values were used.
With respect to Gaylord Gas and Oil, the rate filinc assumes that for the years 1981 to 1983 these units will operate 11 months on gas and 1 month on oil, whereas in the production cost runs gas is assumed for the entire period.
Since oil results in higher fuel costs, the production cost run is conservative relative to the rate filing with respect to delay costs.
H.
CONCLUSION - COST OF DELAY The production cost runs which support the Heins testimony contain two errors in coal costs which produce a non-conservative effect on delay costs. These are the coal costs for the Karn 1 & 2 and the Campbell 3 uni ts. The Licensee indicates that it examined the effects of these coal cost increases and certain other rate case assumptions and found the effects to be negligible.
A production cost run was made on February 16, 1977.
It incorporated the rate case assumoticns on Big Rock out?ut, Campbell 3 and Karn 1 & 2 fuel costs, the retirement of Morrow, and Palisades fuel costs.
Cn a 1981 present worth basis, these adjustments resultad in abcut a 25
, t reduction in delay costs over those reported to the Atomic Safety and Licensing Board in revised Exhibit 14 (Testimony of G.L. Heins).
Based on this analysis, the Licensee concluded that the effects were not significant ar') for this reason, it did not update the Heins testimony in this area.
The Staff has reviewed the February 16, 1977 production cost run and has concluded that the impact of the various adjustments, including reduced coal costs for Campbell 3 and Karn 1 & 2 is minimal.
This computer run did not correct the following differences:
(1) Treatment of Midland capacity (2) Random outage rates (3)
Scheduled maintenance (4) Heat rates (5) Big Rock fuel costs (6)
Fossil fuel costs handling of escalation and round off errors a.
b.
Weadock 1-6, and Karn 3 & 4.
The Staff concludes that adjustments are not required in these instances because: (a) with respect to item (1) it would be inconsistent with the desired purpose of the producticn cost run; (b) with rescect to items (2), (3), (4), and M), the effect would Se negligible; and
.(c) with respect to 6a and Sb, it would only tend to bring the estimated delay costs back towards their initial values.
,. m s3 5.
TREATME'lT OF OEC0!?iISSIO!!I::G COSTS The caterial in tha rate filing pertinent to this issue appears in the direct testimony of Mr. J. S. Ferguson.
Mr. Ferguson estinates the cost for entombment at the end of life, and the annual cost of surveillance, beginning at that time, for all nuclear power plants already in operation on the CP system. Thus, his estimate is applicable to the Palisades and Big Rock Point Nuclear Plants. The value he arrives at is 34", of the total capital cost of these plants. This translates to an estimated cost of $67 million for both units.
However, since these costs are not incurred until the year 2000 for Big Rock Point and 2007 for Palisades, because that is when their operating licenses exoire, this estimate refle. cts the cost in dollars approximately 30 years off into the future.
In the Midland proceeding, the licensee has estimated the cost of deccmissioning the Midland fluclear Power Plant, Units 1 & 2 at S83 million. There is hcwever, one important difference between this estimate and that developed by Ferguson.
The Midland estimate is the 1981 cost of decomissioning in 1981 dollars.
If one wished to compare it with Ferguson's value, one would have to calculate the cost at the end of fiidland's useful life. Aoplying the same escalation rates used by Ferguson (6-1/25 per annum out through 1994 and 5",
per annum thereafter) results in a future cost of about 5460 million in the year 2015.
~-
, This value is significantly larger than the Ferguson estimate and suggests that in the Midland proceeding the deccmmissioning estimate is high relative to what would have been the case had the Ferguson analysis been used. Consequently, in the ?!PC proceeding the decommissioning cost used does not bias the analysis in favor of fiidland.
However, there are other important differences between the two estimates
~
that bear mentioning.
For example, the Midland deccemissioning cost assuces complete dismantling which is a higher level of decommissioning than that assumed by Ferguson. Also, when the Midland estimate is future worthed for comparative purposes, it is future valued between 8 and 15 years more tnan Big Rock Point and Palisades.
Correcting for each of these factors tends to bring the Midland estimate closer in line with the Ferguson value.
II.
Staff Resconse to the Allecations of Dr. Richard J. Timn At the hearings conducted in Chicago the week of May 9, 1977, testimony was presented by Dr. Richard J. Timm which identified the folicwing alleged inconsistencies between the rate filing materials and the Censumer's testimony presented in the Midland prcceeding:
M l '
1.
Differences in Campbell 3 Coal Prices (Midland Intervenors' Exhibits 50 and 51).
2.
Differing Treatment of Capacity Sales of Midland Units 1 and 2 and Campbell 3 -(Midland Intervenors' Exhibits 52 and 53).
3.
Differences in Plant Capacity Factors (Midland Intervenors' Exhibits 54 and 55).
In addition, Dr. Timm alleged at the hearing that Consumers Power Company was forcing bulk purchascs of power in the Midland delay cases and thereby distortiiig delay costs. This contention is also examined in this portion of the report.
1.
DIFFERENCES IN CAMPBELL 3 C0AL PRICES At the hearing, Dr. Timm expressed concern t'o:. the Campbell 3 coal costs were being inflated for delay cost purras 4 r,5995).
On pages 20-21, (Section 4, Item G), the Staff identifis. the discrepancy in the 1984 fuel prices for coal delivered to the Campbell 3 unit and presented the bases for the higher price as reported in the production cost runs.
Furthermore, on pages 22 ?3, (Section 4, Item H), the Staff reported the effect of adjusting for this difference, as well as other discrepancies, and noted that delay costs would be reduced by approximately 25.
%M.
. As reported in Midland'Intervenors' Exhibits 50 and 51 and during cross-examination, Dr. Timm also expressed concern over the 1981 through 1983 coal prices for Campbell 3.
The Staff contends that Dr. Timm is in error in this regard.
First, when Dr. Timm calculated the current dollar Campbell 3 coal costs from the Van Reenen Exhibit, he assumed the real dollar estimates reflected 1977 base values rather than 1975 base values.
- Thus, he ignored an additional year's escalation that sheuld have been added to the Van Reenen values.
Correcting for this results in the rate filing values being elightly higher than those reported for the Midland proceeding whitn if anything would result in lower delay costs in the Midland case than would have occurred had the correct Van Reenen values been used.
Further ore, rather than use the Van Reenen testimony which required certain calculations on Dr. Timm's part, he could have used Mr. Brager's Exhibit JMS-9 frem the rate filing which already reported these values in current dollars. Had Dr. Timm used these data he may have reali:ed that his Van Reenen calculations were in error and that, relative to the rate filing values, the Midland case treatment of Campbell 3 coal prices between 1981-83 is conservative as to their impact on delay costs.
2.
DIFFERING TREATMEMT OF CAPACIT( SALES OF MIDLAND UNITS 1 AND 2 AND CAMPBELL 3 Midland Intervenors' Exhibits 52 and 53 were initially prepared for the rate filing by E. H. Kaiser of Consumers pcwer. Dr. Timm asserts that the trea:-
ment of delaying Camobell 3 on these Exhibits supports his contentien that,
m.,
J in delaying a facility a portion of which has been sold to a third party, it is appropriate to reduce the plant's capability by that portion being sold to the third party. Consequently, Dr. Timm questions the correctness of CP's position that if Midland is delayed, the effect should be judged in terms of the total capacity of these units independent of sales to third parties.
(Tr. 5996).
It is the Staff's position that the purpose of CP's analysis must be fully understood and examined.
CP is attempting to calculate the costs of delaying the Midland Units and in fact these costs would be spread over 100% of Midland's capacity and not just that portion owned by the Licensee.
As explained or. page 13 of this report, (Section 4, Item A), the Staff's view is that the approach taken by Consumers is correct and that wh'at appears to be a discrepancy is a basic philosophical difference in the purpose of the information presented.
Under corss-examination Dr. Timm was asked to ccament on the merits of CP's position as stated on page 4 of its April 13, 1977 letter to the Licensing Board and set cut belcw.
~~
n.. The rate filing is concerned only with the cost of service and rates of charges to Consumers Power custcmers and there-fore looks only at Consumers Power's system; as such, it only considers the portion of jointly owned units which Consumers power is projected to own.
On the other hand, this proceeding must consider the total cost or effect of a potential delay of the Midland units, and,_ as such, must consider their total energy output, not just that portion which Consumers Pcwer is expected to own.
It should be noted that the buy-back provisiens for jointly cwned units are treated the same in both cases.
Consequently, since the studies were made with different objectives in mind, their results cannot be directly compared; Dr. Timm did agree (Tr. 6006-6010) with the fundamental. distinction being made in this quote and thus the Staff believes that the Licensee, the NRC Staff, and the Midland Intervenors' all agree that a difference is warranted, although Dr. Tim believes that the underlying analysis in the Midland proceeding was not adequate.
In surmary, the issue appears not to be the existence of a discrepancy but rather the extent of the analysis required in the Midland proceeding in calculating delay costs.
In the Staff's view, the correct approach is the one taken by the Licensee.
3.
DIFFERENCES IN PLANT CAPACITY FACTORS Dr. Timm testified (Tr. 6019-6020) that substantial differences were evidence in the capacity factors of given generating units for a given year when comparing data developed for the Consumers Power Ccmpany rate O
a n -
r,
.c s3 filing data developed for the Midland hearings.
Dr. Timm did not attempt to suggest w'.ich, if any, of the capacity factors were most representative, but expressed concern that the existence of such differences established that inputs to the different ccmputer models used for the rate filing and Midland proceedings also had to differ considerably.
This comparison was presented by Dr. Timm at the hearing in Midland Intervenors' Exhibits 54 and 55.
The Staff disagrees with Dr. Timm's conclusion that the differences in capacity factors can only be explained in terms of different inputs to the two computer models under consideration. As the first portion of this report makes clear at pages 22-23, (Section 4, Item H), the input differences to the two computer models under consideration are negligible and have an effect of approximately 2% on costs.
The prime reason for the capacity factor differences is a difference in computer modeling.
The rate case model dispatches Consumers' own generating units against only the Consumers' load.
Purchase and Intercharge energy are treated on a historic basis. This type of approach has been consistently used in Consumer's rate applications to the Michigan Public Service Commission.
Such an approach would not be correct for the delay costs analysis presented by Consumers in this proceeding.
To accurately develop those costs, a ecmputer model is required which dispatches the ccmbined Michigan Electric Coordinated System (MECS) generation against the MECS load.
s,-
.s_
31-This is what Consumers has actually done in performing its delay. cost calculations with its production cost program.
The capacity factors for the Consumers facilities should decrease as low cost generation available from other MECS members would displace the higher cost generation of those units.
This is the behavior seen on Midland Intervenors' Exhibit No. 55 with two exceptions.
First, Campbell Units 1 and 2 have increased capacity factors.
This is explainable as these are Consumers' lowest cost units and would be run harder to supply the combined MECS load.
Second, the Morrow 1-4 units have a sharp increase in capacity factor.
This is due to the fact that in the rate case, Weadock Units.1 through 6 and Karn Units 3 and 4 are run narder than the Morrow Units since the Karn and Weadock Caits are subject to a "take or pay" clause in their No. 6. fuel oil contract.
Failing to run these units would produce '
econcmic penalties so they are run in the rate case model thereby lowering the Morrow capacity factors.
The "take or pay" provision expires in 1981 and so would not influence the 1984 production cost runs which shcw a correspondingly higher capacity factor for the Morrow Units.E/
S/ The modeling of the "take or pay" provision was not changed in the rate case program for 1984 as that year, as well as all years beyond 1980 are of no consequence in a rate case application.
The mcdeling in the production cost run for the Midland delay cost calculations is correct.
~ -'
s
. One other difference on Midland Intervenors' Exhibit No. 55 requires explanation and this is the capacity factor related to Big Rock.
The capability rating of Big Rock was assumed to be 61 MN in the rate cas: study as compared to the 71 FM rating used in the tiidland hearings.
The assumed energy produced in the rate case study in 1984 was 109,000 Mwbrs. as ccmpared to 444,027 Mwbrs. used in the Midland study.
The Staff has calculated the capacity factors for Big Rock under the two different assumptions and found the capacity factors set forth in Column's C and 0 to be correct for the assumptions made.
So the Big Rock capacity factor difference is readily explainable as related to differences in generation and rating. These differences are negligible in terms of their cost impact as was elaborated en in the first portion of this report.
In summary, the differences in capacity factors are mainly attributable to the use of two different computer ecdels. The producticn cost cceputer medel was the correct model to apply for the Midland delay costs.
The rate case model is the one which has been historically used in Consumers' rate filings.
The Morrow 1-4 and Big Rock capacity factors cannot be related to model differences.
Rather different input assumotions were made which are readily explainable and have negligible impact en delay Costs.
6 e
w
_. FORCED PURCHASES CONTENTION Dr. Tim alleged at' the hearing that Consumers Power was forcing purchases of bulk power in their production cost runs thereby distorting the costs of replacement power associated with a delay of the Midland Units.
(Tr.
5998-6004).
In his testimony, Dr. Timm referred to a production cost computer run numbered C4-043 which did not include the " Purchase 10" category.1E/ In Dr. Timm's view, the lack of any " Purchase -10" in this computer run supports-his argument that Consumers inappropriately included
" Purchase 10" in the delay cases analyzed in this proceeding.
The Staff has reviewed the allegations of Dr. Timm and concludes that they are unfounded.
The particular computer run referred to by Dr. Timm, C4-043, was only one of a large number of production cost computer.run; used by Consumers Power to conduct sensitivity studies.
Its use would not have been appropriate in examining the Midland delay cases.
- Rather, that run and others would have been used to determine the appropriate level of " Purchase 10" for use in the Midland delay cases.
3SI " Purchase 10" refers to bulk power purchases which Consumers Power schedules to maintain a 20% installed reserve margin in the event that the Midland units are delayed.
.<n
%s
. A production cost computer program attempts to model the actual operation of a power system under given assumed conditions of: demand, energy for load, generation availability, purchase power options, incremental heat rates, fuel costs, fuel availability, contractual constraints, regulatory constraints, and transmission limitations.
At present, no ccmputer model has been developed that can duplicate the actual operation of a power system under this series of given conditicns.
Despite the limitations of modeling, the production cost ccmputer programs available to the utility industry today contribute greatly to the decision-making process. Various techniques are used to compensate for the limitations.
The iteration method is one such technique used by the utility industry to tailor the computer outputs of specific parameters, such as " Purchase 10" power. By this method, an assumed capacity factor is assigned to the firm capacity purchase to establish the impact, if any, on the output of generating units having known costs that are icwer than the projected purchased energy cost.
If the results show a reduction in such low cost generation, another capacity factor is assigned in an iterativ e process until the proper capacity factor for the purchased firm capacity is established.
With good judgement and experience, the proper value can be established in
-a limited number of iterations.
-.. This is the procedure that Consumers Power Company used to determine the proper capacity factor to assign to the " Purchase 10" power required in the flidland delay case production cost computer runs. Computer runs were made using three different capacity factors for the " Purchase 10" power.
A zero percent capacity factor case was run to establish the maximum generation that could be expected from the low cost generating units.
A 90 percent cpacity factor was then used to establish the sensitivity and provide a base from which interpolation could minimize the number of interations required to establish the proper value.
A 70 percent capacity factor was found to be the proper assignment since the. impact on the low cost units was negligible.
The use of a 70 percent capacity factor by Consumers Power t.o develop the quantities of " Purchase 10" required was reasonable and did not result in forcing any unnecessary purchases.
This can be verified by examining the several computer runs used in the iterative process to ' develop the 70 percent figure.
-e-
-v,, -
.+
.., The follow-ng tabulation details the 1982 reducticn in generation from relatively lower cost units when the " Purchase 10" power is assigned a 70%
capacity factor as compared to a zero percent capacity factor: (Listing lowest cost units first).
Unit Total Generation "Whrs x 1,000 Generatina Unit Ra tina-MW Zero%
70%
Decrease Campbell #1 267 1,845 1,845 0
Campbell'#2 372 2,395 2,395 0
Waaock #3 159 962 957 5
Weadeck #7 159 997 993 4
Cobb #5 167 1,072 1,072 0
Cobb #4 158 969 968 1
Whiting #1 106 590 585 5
Whiting #3 133 792 708 a
Whiting #2 106 676 671 5
Campbell #3 7a2 4,579 4,5a8 31 Cobb #2 68 347 343 4
Cobb #3 68 347 341 6
Cobb #1 68 345 340 5
Karn #2 264 1,641 1,618 23 Karn #1 265 1,458 1,dal 17 Total Reduction =
110
.... The above data is extracted from the 12/81 delay case where 190 lM of
" Purchase 10" pcwer was required to maintain the 20", reserve margin.
Tile total energy purchased under the " Purchase 10" demand of 190 MA for 1982 was 896,216 Mahrs.
The 110,000 Mwbr. cutback on the relativelv low cost units represents approximately 12.3 percent of the energy purchased.
Standing alone, the above figures may suggest that a lower capacity factor than 70", could be justified to avoid forced purchases.
- However, other elements must be factored into the ultimate conclusion.
For example, when " Purchase 10" power is available during the peak hours (as it was),
~
there is less need for pumped storage capacity to meet the peak demands.
It follows that less pump'ng pc.er will be required during the off peak
' hours to provide this peaking energy and a corresoonding reduction in generation frcm the icwer cost unit results.
Generating units that are more competitive (economically) with " Purchase 10" energy at a given hcur, may not be ccmpetitive when the substitution must go through the pumping process.
The penalty factor for a gh en unit t
to provide pumping pcwer is about 1.4.
For example, assuming purchased pcwer at 20 mills per kilowatt hour, a generator with an incremental producticn cost of 18 mills /Xwhr is the econcmic choice between the two sources. Mcwever, if the energy must be generated off-peak to supoly purping pcwer to be used the following day on-peak, then the generation will cost 18-x 1.4 or 25.2 mills / kwhr ccmpared tr " Purchase 10" at 20 mills /kwnr.
~
s
. Naturally, the " Purchase 10" power is cheaper and should be used even though the initial cost of the generation is only 18 mills / kwhr.
In su= mary, the Staff does not concur with Dr. Timm's conclusion that Consumers. Power Ccmpany was forcing " Purchase 10" power. Rather, Consumers assigned reasonable values to the bulk power capacity required as well as the capacity factor assigned, and thus the delay costs were properly calculated.
Conclusion The Staff has found no discrepancies between the rate filing and the Consumers Power testimony to warrant further consideration by this Board.
In general, the allegations made by Dr. Timm are either unfounded or constitute a difference of opinion as to the appropriate acdel to use for calculating delay costs.
In certain instances, the Staff has detected inconsistencies and errors in the materials reviewed. These inconsistencies and errors appear random in nature and do not reflect any intent on the part of the Licensee to introduce a systematic bias.
The impact of these itens on the conclusions presented by the Licensee to this Board is minimal.
( D
-~,
ATTACHMENT B UNITED STATES OF AMERICA f
!;UCLEAR REGULATORY COMMISSION i
BEFORE THE ATOMIC SAFETY AND LICENSIflG BOARD In the Matter of
)
)
CONSUMERS POWER CCMPANY
)
Docket flos. 50-329
)
50-330 (Midland Plants, Units 1 and 2)
)
AFFIDAVIT OF SIDMEY E. FELD Sidney E. Feld deposes and says under oath as follows:
1.
I am a Regional / Environmental Economist in the Division of Site Safety and Environmental Analysis, U.S. Nuclear Regulatory Ccmmission. My professional qualifications were admitted into evidence in this proceeding on February 16, 1977 following Tr.
2.
I assisted in the preparation of the " Report Prepared by'S. Feld and W. Gundersen for the Atomic Safety and Licensing Board on Alleged Discrepancies Benteen Consumers Pcwer's Rate Filing of January 31, 1977 and Testimony Presented to the Atomic Safety and Licensing Board in the Midland Proceeding" consisting of 38 pages.
I hereby certify that this report is true and correct to the best of my knowledge and belief.
/
l' 9
- b. N/)
.p q
Sidney E. Feld Subscribed and sworn to before me tnis m % day of May, 1977.
bdRAAd NotaryPuoliQ My Commission expires: QL \\ ; 6%.
\\i
ATTACHMENT C s l UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSIOM BEFORE THE ATOMIC SAFETY AND LICENSING BOARD i
In the Matter of
)
CONSUMERS POWER COMPANY Docket flos. 50-329 50-330 (Mi'. land Plants, Units 1 and 2)
)
AFFIDAVIT OF WALTER J. GU!!DERSEN Walter J. Gundersen deposes and says under oath as follows:
1.
I am employed by the Federal Power Cocmission as Assistant ta the Chief of Division of Power Supply and Reliability. My professional qualifications are attached to my testimony and were admitted into evidence in this proceeding on March 24, 1977 following Tr. 5101.
2.
I assisted in the preparation of the " Report Frepared by-S. Feld and W. Gundersen for the Atomic Safety and Licensing Board on Alleged Discrepancies Between Consumers Power's Rate Filing of January 31, 1977 and Testimony Presented to the Atomic Safety and Licensing Soard in the Midland Proceeding" consisting of 38 pages.
I hereby certify that this report is true and correct to the best of my kncwledge and belief.
,'! ~ ~ ~
,3 Walter J. Guncersen Subscribed and sworn to before me this g day of May, 1977.
Om1 Q. xAa Notary ?uolicq My Commission expires: A.L 1. R'N.
g L'