ML20005D908

Forwards Response to Request for Addl Info,Amend 9 to SAR & Amend 11 to Tech Specs in Support of Pulstar Reactor Relicensing Effort,Per 890920 Request

ML20005D908
Person / Time
Site:	North Carolina State University
Issue date:	12/18/1989
From:	Monteith L North Carolina State University, RALEIGH, NC
To:	Alexander Adams NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
Shared Package
ML20005D909	List: ML20005D908 ML20005D915 ML20005E608
References
NUDOCS 9001020234
Download: ML20005D908 (52)

Text

.

8:

i North Carolina State University i

Offiee of the Chancellor

= Box 7(01 Raleigh 270957001 Deeembor 18, 1989 (910) 737 4191 Document Control Desk Attention: Alexander-Adams, Jr.

U. S. Nuclear Regulatory Commission Washington, D.C.

20555 Re:

License No. R-120 Docket No. 50-297

Dear Sir:

In accordance with the transmittal dated September 20,

~1989, from Mr. Alexander Adams, Jr., Project Manager, to Dr.

Bruce Poulton, then Chancellor of North Carolina State University, and the follow-up transmittal dated October 31, 1988, from Mr. Alexander Adams, Jr. to Dr. Thomas S.

Elleman, Nuclear Engineering Department Head, the following documents are officially submitted:

(1)

Our response to the " Request for Additional Information" questions and associated attachments.

(2)

Revised Safety Analysis Report (SAR) sections, Xmendment 9 that were specifically modified in support of our response to the " Request for Additional Information."

The attached sections of the PULSTAR Safety Analysis Volumes I and 11 supercede and replace all text pages previously found in the PULSTAR SAR.

All figures submitted under the existing Amendment 8 remain current, and therefore, were not resubmitted under this Amendment 9 transmittal.

(3)

Revised Technical Specifications. Amendment 11, as requested by question #72 of the " Request for Additional Information."

Amendment 11 of the Technical Specifications supercedes and totally replaces previous versions.

All of the above documentation has been prepared by the PULSTAR staff in support of the PULSTAR Reactor re-licensing effort.

In regards to the PULSTAR Safety Analysis Report and Technical Specifications, sections that have significant changes have been marked with revision bars.

Simple grammatical / nomenclature corrections and format changes are not marked.

9001020234 89121G 4010 PDR ADOCK 05000297 P

PDC fl North Carolina State University is a land grant university and a constituent instituttor of The University of North Carolina.

i

c U. S. Nuclear Regulatory Commission Page 2 December 17, 1989 i

If there are any questions on this matter, contact Mr.

Garry D. Miller, Associate Director, Nuclear Reactot Program, at (919)737-2321.

Sincerely, e )hN f cm Larry K. Monteith Interim Chancellor cc:

Garry D. Miller, Associate Director Thomas C. Bray, Reactor Operations Manager William D. Morgan, Radiation Protection Officer Richard Mowat, Chairman, Radiation Protection Council Hayne Palmour Chairman, Reactor Safeguards Advisory Group E.

J. McAlpine Division of Radiation Safety and Safeguards U.

S. Nuclear Regulatory Commission Region II 101 Marietta Street, N.W.,

Suite 3100 Atlanta, Georgia 30303

?

Notary Public My commission expires July 8, 1991.

North Carolina State University Nuclear Reactor Program PULSTAR Reactor License # P.120 Docket # $0 297 Response to Request for AdditionalInformation Revision 0, December 15, 1989 PULSTAR SAFETY ANALYSIS REPORT 1.

QUESTION: Section 1.3.2 Licensees must comply with 10 CFR Part 20, it is not a guide. Furthermore, current practice is also to implement an ALARA program, and you are expected to address that in all radiological considerations.

Please adress this comment.

ANSWER: We acknowledge and operate under the fact that 10 CFR 20 is the defacto law concerning our radiological activities. Amendment 9 of SAR Section 1.3.2 removes the word " guidelines values

• and replaces it with

• regulatory limits".

In addition, we fully support the concept of ALARA (Refer to the response to SAR question # 55).

2.

QUESTION: Please clarify in your SAR what design features you have considered to be engineered safety f-atures.

ANSWER: Using the definition on an " Engineered Safety Feature (ESP)" as

• those features or systems that are present to mitigate the potential consequences of accidents", the ventilation system, particularly the confinement portion of the system (charcoal and HEPA Filters, Confinement Fans, associated dampers /ductwork, supporting instrumentation, and auxiliary power) would be considered an ESF.

3.

QUESTION: Section 1.4.4 - What are the abnormal conditions referred to in this criterion and how is control rod insertion accomplished.

l l'

ANSWERt An abnormal condition exists when the control rod fails to fall l

completely into the core upon a SCRAM, such as if the control rod Wde was to

• stick" or " bind" due to excess friction. The SCRAM circuitry both turns off magnet current, allowing the control rods to fall by gravity, and initiates a

" Reverse" or "Run back" whereby the control rod drive mechanisms start driving toward their fullin position. The drive mechanisms would function to push the control rod blade in, overcoming the friction that caused the sticking / binding.

As a backup to the SCRAM circuitry, the reactor operator can also manually drive the control rod drive mechanisms into the core to assist in fully inserting the control rod blade.

4.

QUESTION: Section 1.4.7, Criterion 7 10 CFR Part 100 is not applicable to research reactor.c. Because there is no comparable regulation for research reactors, NRC expects you to compare with 10 CFR Part 20, and the ICRP l

1

NCSU PULSTAR December 15,1C9 Resposee to Request for Additional leforniation Revision 0 recommendations on which it is based, with AIARA considerations. Please address this in all applicable parts of your SAR.

ANSWER: We concur and consider 10 CFR 20 the law for radiological actisities. In addition the pertinent ICRP recommendations such, as ICRP 26 dealing with AIARA, are also applicable to our facility.

. 5.

QUESTION: Section 1.4.11, Criterion 11 - You are requested to discuss in quantitative terms, in the appropriate sections of your SAR just what is monitored, where, by what intruments, etc. Discuss instrument efficiency, ranges, calibration methods and frequencies. Discuss radiation monitoring in event of accidents.

ANSWERt Figure 101 details the radiological monitoring at the PULSTAR Facility indicating type of detector and location. SAR Sections 5.2.2 and 10.2.2 discuss the function of the various ' radiation monitors. The attached Table 71 of the PULSTAR O)erations Manual details actual setpoints and automatic functions that are in tiated by the various setpoints. The following radionuclide sensitivities for the stack monitors was reproduced from Appendix 4.1 of the PUI. STAR Emergency Procedure 4.0,

• Emergency Classification *:

Stack Gas

• Ar 1.04 x 10' cpm / Ci/cm' "Kr 3.9 x 10' cpm / Ci/cm'

• Xe 5.59 x 10' cpm / Ci/cm' Stack Particulate

Ar 1.28 x 10" cpm / Ci/cm' "Kr 7.23 x 10" cpm /pCi/cm' "Xe 4.075 x 10" cpm /pCi/cm' Auxiliarv GM

Ar

~ 4.0 x 10' cpm / Ci/cm' "Kr 1.2 x 10' cpm / Ci/cm' "Xe 4.7 x 10' cpm /pCi/cm' The area monitors are Victoreen 845 Systems using an ion chamber with a range of 0.1 to 10' mR/hr. Calibration of the stack gaseous detectors is performed on an annual basis (as required by Technical Specification 4.4) and is accomplished by a combination of electronic calibration and a two source verification of detector response. The Stack Gas channel also receives a primary sensitivity calibration by injecting a known quantity of Ar into the system every 5 years. Calibration of the area monitors is accomplished annually by placing the detector in a Victorcen field calibration unit, Model 848 8, and thus verifying the measured dose rate is correct. During an emergency, it is anticipated that the channels would continue to operate, on auxiliary power if necessary, to monitor both area doserates and effluent activities. Since the readouts are in the Control Room and therefore segregated from the Reactor Bay air space, access to the readouts is available during the emergency.

2

NCSL) PUIETAR Deceniber 18,1989 Response to Request for Additional leforniation Revision 0

{

6.

QUESTION: Table 1.1 - Table 1.1 indicates some inconsistencies: e.g. on page 5 you say that only a 5 x 5 array of fuel elements is considered, but the Table 1.1 footnote says a *5 x 4 Standard Core." However, the number of pins is 625.

There are several other inconsistencies; e.g. WNYNRC is no longer its name, height of pellets of BMRC is wrong, number of pins in BMRC is not consistent with fuel element dimensions, H,O/UO, ratios seem inconsistent, capture to-fission Ialt, perhaps others. Please edit and correct.

t ANSWER: Amendment 9 of SAR Table 11 includes the requested revisions.

7.

QUESTIONt section 2.1.3.1, 2.1.3.2 - Please discuss (or reference) in more detail the directions and distances from the reactor to the nearest permanent i

residences, the nearest occupied campus building, the nearest occupied student housing, etc., and the quantitative radiological implications. What is the approximate population lof Raleigh,and the Raleigh metropolitan area?

ANSWER: The nearest campus building is Polk Hall, located approximately 33 meters west of the Reactor Building. The nearest student housing is Carroll Hall located approximately 260 meters west southwest of the reactor. The nearest permanent residence is in the vicinity of the Wachovia Bank located approximately 240 meters northwest of the reactor. The nearest " Emergency Site Boundary" is West Broughton Street, located 12 meters from the Reactor Building. In the case of the maximum credible accident (complete loss of pool water), the calculated dose rate at the site boundary is < 0.3 mrem /hr.

Airborne releases associated with normal *Ar oroduction is detailed in our response to SAR question # 57. Airborne releases asociated with fuel failure l

are detailed in our response to SAR question # 63. Raleigh has an L

approximate population of 200,000 and the greater metropolitan Raleigh area l

wh!ch includes the cities Cary and Garner has an approximate population of 330,000.

8.

QUESTION: What is the location (direction and distance) of the closest in use railroad line to the facility? What volume of rail traffic does this line carry?

What plans does the University have in place to respond to a railroad accident?

To what extent would these plans involve the reactor facility?

ANSWER: The Southern Railway traverses the NCSU campus 377 feet south of the reactor facility at the closest point. Average rail traffic year round is one 50 car freight and one 8 car passenger train daily. NCSU would rely on the Wake County Emergency Management contingency plan, should a railroad accident occur in proximity to the reactor facility. Notification to Wake County of a railroad accident on the campus would be made by NCSU Public Safety and NCSU would regulate pedestrian and vehicular traffic at the accident site, assisted by the City of Raleigh and Wake County personnel. It should be noted that there exists a row of large multi-story university buildings between the railroad and the reactor facility, that would serve to buffer the PULSTAR Reactor in a railway accident, such as a derailment. During a railway emergency, the reactor staff would ensure the timely evacuation of the reactor facility (if required by the emergency), but would not serve an active role in directing recovery activities unless the PULSTAR Reactor was directly affected.

3

NCSU PUt3 TAR Deceanber t$,1989 Response le Roguest fee Additional 1 formation Revisloe 0 9.

QUESTION: Please provide information on the reactor room wall thickness and details of roof design. How many penetrations does the reactor room have and how is air leakage through these penetrations controlled?

ANSWER: The reactor room (Reactor Bay) wall thickness is sixteen inches, composed of a twelve inch steel reinforced concrete layer and a four inch brick veneer. 'Ibe Reactor Bay roof is a three inch steel.relnforced concrete layer with 6" x 18" steel reinforced monolithic concrete webs on 30" centers. In addition, three 16" x 33" steel reinforced concrete joists span the Bay, centered over the reactor biological shield.

t There are 127 penetrations in the Reactor Building walls which lead to outside ambient atmospheric pressure. Because the Reactor Bay, Control Room and Mechanical Equipment Room (MER) are maintained at approximately the same negative air pressure (with respect to outside atmospheric pressure), all penetrations are considered. Air leakage through the penentrations, i.e., bypass flow adjacent to pipes, conduits, etc., due to imperfect fit is eliminated by poured hydraulic cement and/or grout as a sealing agent. Silicone RTV is appplied to very small pathways not scaled satisfactorily by cement or grout.

10.

QUESTION: All of Chapter 2 Please discuss the quantitative radiological implications of this chapter, and why there is deemed to be no substantial reason that the reactor location does not continue to be acceptable. If you take credit for building characteristics not yet discussed in detail in your SAR, give cross references to the appropriate sections, e.g. the consequences if an earthquake were to occur.

ANSWER: Since the NCSU PULSTAR was first constructed and placed into operation in 1972, only minor changes in the immediate campus vicinity have transpired. These would include the recent addition to Daniel Hall, and the D.H. Hill Library renovation. None of the building changes affect analysis presented in the SAR. In addition, very few changes have transpired in the local residential areas, with no new permanent housing located in the reactor vicinity since the original startup Ground transportation routes also remain significantly unchanged during the original operating period. The local RDU Airport has experienced significant growth during the original operating period, l

particularly with the addition of the American Airlines hub. However, the airport is approximately 10 miles from the reactor site with its main runways arranged such that landings and takeoffs are not in the immediate path of the reactor. Therefore, the growth of the local airport does not affect presions site acceptability. Catchment basins and drinking water reservoirs remain at favorable locations with respect to the reactor facility. Meteorological conditions of the reactor site continue to be acceptable with little variation observed over the first operating period. No new major faults have been discovered near the reactor site nor significant earthquake activity has transpired in the first operating period. In summary, the site continues to represent an acceptable location in terms of meteorological, seismic, and commercial / residential development conditions.

11.

QUESTION: Section 2.2 - Please describe your method for obtaining meteorological data in the event of an accident that involved releases. If this 4

NCSU PtMTAR December t$,1989 Response to Request for Additional Information Revision 0 involved interaction with an agency outside of the University, letters of agreement should be provided.

ANSWER: The Division of Emergency Management (within the North Carolina Department of Crime Control r.nd Public Safety) is the off site organization that would provide meteorological data upon request during an actual emergency.

This group has dedicated phone lines to the National Weather Service office located at the Raleigh Durham International Airport. The letter of agreement for this organization is found in Appendix A of the PULSTAR Emergency Plan (and is presently on file with the Commission).

12. QUESTION: Section 2.2 - Please provide updated meteorological data, for example, section 2.2.1.3 discusses the time period 1916 to 1958.

ANSWER: Table 212.of SAR Se.ction 2 details temperature, humidity, precipitation and wind data for the period of January 1983 to December 1987 and is summarized by month. Table 2-13 of SAR Section 2 details time of day averages for resultant wind speed / direction, average speed, and the number of precipitation events for the period of January 1983 to December 1987. These tables provide typical meteorological information from the latest published NOAA reports in the Raleigh area and supplement the older historical periods I

also detailed in the SAR. At the time of submittal of Amendment 8 of the SAR in August 1988, the high and violent wind information (Section 2.2.1.3) was still representative of this area. For example, the highest sustained winds l

recorded were in 1954 from Hurricane Hazel (73 miles / hour) with a NNE l

direction and occurred during the month of October. In 1962 during the month of July, sustained winds of 69 miles / hour (SW direction) were measured at the Raleigh Durham International Airport. Later in May 1972, sustained winds of 54 miles / hour (SW direction) were recorded. However, Raleigh was i, truck by one or more tornadoes on November 28,1988 resulting in $ 77 million in l

property damage,4 deaths and 157 injuries within the state of North Carolina.

At its closest point, a tornado passed within approximately 5.5 miles of the PULSTAR Reactor facility. The track of the Raleigh tornado was 84 miles long and was estimated at its onset as F4 severity on the Fujita Scale with winds estimated at 210 miles / hour. This particular tornado had an estimated ground speed of 50 miles / hour and passed through Wake, Franklin, Nash, Halifax and Northampton Counties. Prior to this date, only twelve tornadoes had been reported within the state of North Carolina during November for the period of 1916 to 1987. For the same period only eight tornadoes were reported in December. In the 72 year recorded history of tornadoes, only one tornado was reported in Wake County in both November (1%6) and December (1942). On the average, North Carolina experiences approximately 12 tornadoes per year.

During the period of 1884 to 1984, only six F4 severity class tornadoes have been recorded in North Carolina.

13. QUESTION: Section 2.3.2 - Discuss drainage between the reactor site and i

1.akes Raleigh, Johnson, and Benson.

l ANSWER: Surface drainage from the immediate reactor site does not drain to any of the three referenced lakes (Raleigh, Johnson, or Benson). Site drainage flows to the Rocky Creek which then flows into Walnut creek about 2.25 miles i

I 5

NCSU PULETAR Dember 15,1%9 Response to Request for Additlemal leforunatlos R nisloa 0 downstream of Lake Raleigh. Walnut creek ultimately flows into the Neuse River which flows into the Atlantic Ocean. Lakes Johnson and Benson are on 1

different catchment basins that do not include the reactor site.

14.

QUESTION: Section 3.1.1 De pulsing achieved doesn't seem consistent with information in Table 1.1. Because of possible implications to inadvertent transient accidents, please discuss. One of the objectives of an FSAR is to compare actual observations with initially projected characteristics. Please do that.

t ANSWER: The highest peak power pulse achieved for the NCSU PULSTAR was 980 MW while it was originally designed for 2200 MW. The Pulse Rod ejection time is primarily controlled by the accumulator pressure setting, referred to as the " pulse pressure". The PULSTAR could have operated with up to 65 psig pulse p)ressure, however, during shutdown pulse ejection testing (foll construction it was concluded that a lower pressure value of 45 psig would be desirable (reference PULSTAR Startup Test 2.14). This decision was made to reduce the potential for physical damage to the bridge superstructure and components mounted on it due to the abrupt Pelse Rod piston operation. The difference in expected versus actual peak power is a direct result of the reactivity insertion being slower allowing the fuel temperature coefficient to terminate the pulse at a lower peak power.

15.

QUESTION: Section 3.1.2.1 - (a) Please analyze the adequacy and accuracy of a shutdown margin of.004 delta k/k. (b) Now that the Pulse Mode is no longer to be used, the role of the so-called Pulse Rod is changed. Explain its role in the definition of shutdown margin. Justify the steps taken to " disable pulsing capability."

ANSWER:

(a)

In order to fully resolve questions concerning the 400 pcm shutdown margin limitation, we are proposing an upper limit on the excess reactivity of 4070 pcm. The operational requirements for the PULSTAR are typically as follows:

Xenon (for typical o'perations) 600 pcm Temperature coefficient 140 pcm Power Defect (to 1 MW) 330 pcm Total 1070 pcm If we add an additional 3000 pcm allowance for experiments, the total excess reactivity requirements yield 4070 pcm, hence our proposed 4070 pcm limit. The fundamental criterion for an upper limit on excess reactivity is maintaining ensured capability to shut down the reactor, hence the minimum shutdown margin. The limit on excess reactivity also ensures that the SAR analyses are applicable to the operational core. As for the 400 pcm shutdown margin adequacy, the 5 x 5 Reflected Core #3 Total Gang worth is 8310 pcm with the highest worth rod (Regulating Rod) equal to 3770 pcm. With the maximum allowed excess reactivity of 4070 pcm, the shutdown margin would calculate to be 8310 3770 - 4070 = 470 pem, therefore, adequate shutdown margin is maintained. In terms of the 6

NCSU PUtATAR Dea mber 15,1989 Response le Request for Additional Informatloa h ion 0 accuracy of excess reactivity measurements and rod worths, based on our l

experience we can easily determine these values within 50 pcm. Technical Specification Section 3.2 has been revised to include the excess reactivity limit which will ensure that the required shutdown margin is maintained.

(b) No shutdown or scramming credit is allowed for the Pulse Rod during calculation of the shutdown margin. The measured excess reactivity for the shutdown margin calculation is taken with the Pulse Rod fully out. Since much of the original hardware for the Pulse Rod will remain and it does not have SCRAM capability as does the remaining control rods, we chose to continue to refer to this rod as the " Pulse Rod" or " Shim Rod". This later name is based on the fact that its primary purpose is to now to support calibration of the remaining scrammable control rods. To disable the ejection capability of the Pulse Rod, the air line supplying the piston aressure is to be removed and permanently capped at the Pulse Rod.

Removal of the air supply precludes operation of the Pulse Rod piston mechanism. Note that the standard dnve features, i.e. 7.5 inches / minute travel by lead screw rotation, will still exist.

16. QUESTION: Section 3.2.1

' Vacant core positions" are mentioned. Please be more specific and provide the analyses necessary to justify such use. Include in the analysis the accidental dropping of a fuel assembly into the highest worth vacant core position possible.

ANSWER: Operation with a core arrangement with a vacant fuel assembly position would require considerable startup testing, particularly if the core had less then 25 fuel assemblies. However, such an arrangement is feasible and could meet the various reactivity thermal hydraulic criteria. Technical Specification 4.2 provides the reactivity measurement requirement for a new core configuration. The revised Technical Specification 5.1 provides the appropriate flux mapping requirement for a new configuration. The SAR as written does not address the reactivity and thermal hydraulic behavior of a core operated at power with a vacant fuel position and therefore, the mentioning of a vacant aosition was included as a potential future (onsideration. In order to clarify this issue, Amendment 9 of SAR Section 3.2.1 now states " Core access is available by direct insertion of samples through'the pool water into flooded or dry exposure ports on the core periphery."

17.

QUESTION: Section 3.2.1 - Please provide discussion of the effects of the following: heat capacity, thermal diffusivity, retention of fission products,

" benefits" of sintered pellets, physical supports that prevent bowing, relative core sizes of SPERT and NCSU PUIJSTAR (or relative power and energy densities).

ANSWER: The main advantages experienced from UO, fuel instead of metallic uranium fuels, are (a) higher permissible fuel and plant operating temperatures, due to much higher melting points; (b) good irradiation (dimensional, structural and volumetric) stability, due to the absence of low temperature phase transformation; and (c) high corrosion resistance to oxidation and coolant attack, as a result of the relative inertness and compatibility with the cladding and coolant in the reactor. The accepted theoretical density of UO, is 10.96 g/cm' and the sintering process is used to fabricate the pellets in such a manner that 7

f l

NCSU PULSTAR Denmber 18,1989 nnponse to Requot for Additional laformation Re m los 0 the actual density of the pellet is as high as possible. A high

• actual" density of UO, has the following important advantages: (a) high uranium atom density, (b) large thermal conductivity for heat conduction, (c) high capability to contain and retain fission product gases in the fuel and (d) large linear power rating of the fuel assembly (reference. Nuclear Reactor Materials and Applications by Benjamin M. Ma,1983). When operated at high tem >cratures with a large thermal gradient such as in a power reactor, low dens ty porous fuels can form a central void in the axial direction in its early operating life. 'niis phenomenon results from the migration of fabricatec. P.s up with the thermal gradients by a vapor transport mechanism to the centerline of the fuel pin. High density UO,,

in conjunction with low operating temperatures / temperature gradients (across the fuel pellet radius) such as exist at the PULSTAR, reduces the potential for this phenomenon. The as delivered PULSTAR fuel had actual densities ranging from 10.51 to 10.76 g/cm' (or 95.8% to 98.1% of theoretical density). As a

heat capacity of the UO, fuel allow,ing tests at BMRC (Reference 31), the demonstrated by the prototype puls s for pulsing to high levels of energy releases j

without fuel damage. The effect of the heat capacity can be seen in the calculation of total fuel temperature rise during a pulse as shown in Appendix 3C. The thermal diffushity comes into play during a pulse in slowing down the energy transfer from the fuel pellet to the water surrounding the fuel rod.

Again, BMRC pulsing tests determined that the onset of film boiling does not occur until relatively high levels of energy release as a result of the slow heat transfer from the fuel. In addition, an explosive formation of steam (that might be experienced with metallic plate type fuel) is avoided during the pulse (and immediately following the pulse) as a result of the low thermal diffusivity. The favorable retention of fission products appears under the Section 13 analysis for fuel pin clad failure and demonstrates that the high retention values for UO, i

result in low offsite consequences for a postulated fuel pin failure.

Based on experience in the SPERT pulsing test, the PULSTAR fuel is outfitted with spacer warts that maintain the fuel rod axial positions along its length.

i Figure 3-4 depicts these warts that prevent the rod from bowing during pulsing l

(or an accidental transient). The warts are placed at eight inch intervals on the j

PULSTAR fuel. The SPERT fuel had an active length of 66.9 inches while the l

PULSTAR has 24.0 inches.

I

18. QUESTION: Section 3.2.2.2.

Explain why a failure of the Pneumatic System cannot syphon water out of the pool.

ANSWER: A siphon break hole exists in the highest point of the pneumatic system to prevent siphoning. In addition, the air leakage in the blower system at the 417 feet level will break any siphon.

19. QUESTION: Section 3.2.2.5 - (a) Please give more justification for gang withdrawal of control rods. It is more conservative and realistic to assume the maximum rate of reactivity insertion not the average permitted by technical i

I specifications. (b) Are there credible actions by which withdrawal speeds might increase?

ANSWER:

(a) The maximum reactivity addition rate associated with the gang rod 8

NCSU FULATAR Deceanber 18,1989 Response D Request for Additional leforenation Revkloa 0 withdrawal for the 5 x 5 Reflected Core #3 is less than 100 pcm/sec. The Amendment 11 of the PULSTAR Technnical Specification Section 3.2 now states that the 100 pem/sec applies to the maximum reactivity rate in the critical region and is not averaged over the length of the rods.

(b) No, since the control rod drive lead screw is driven by a synchronous motor whose speed is governed by the 60 Hz frequency of the power supply.

20. QUESTION: Section 3.2.0.6 Please discuss the design and use of irradiation brackets.

ANSWER: ne typical irradiation " basket" utilized in the expcsure ports is made of 6061 T6 aluminum. The basket has many perforations to allow water to freely flow around the sample containers to remove any heat production from gamma heating. The baskets employ a

• clam-shell" lid that allows the sample bottles to be removed from the irradiation basket underwater remotely if required. The baskets are lowered into the exposure port by a nylon string. To reduce buoyancy when filled with sample bottles, the baskets contain a small lead weight encased in aluminum. Samples are generally placed in a heat sealed vial and then positioned with other samples inside a screw top poly bottle. The frradiation baskets can contain up to 7 vertically stacked poly bottles. By virtue of the Rotating Exposure Ports (REP), the samples (including irradiation basket) can be rotated to provide a more uniform fluence among the individual samples l

l 21.

QUESTION: Section 3.2.27 - Please provide additional explanation of the 75 MWsec energy release statement.

ANSWER: The 75 MWsee statement was based on a physically unrealistic, but an ultra conservative scenario to estimate the maximum annull temperature (which could then be used to estimate the annull pressure). The scenario involved the assumption that a fuel rod is totally insulated and a pulse causes an adiabatic and uniform rise in the average fuel rod temperature up to the melting point of zircaloy (3365'F). Therefore, one could refer to the 75 MWsee as a i

• clad melt" number. This terminology has its origin in the AMF Advanced Pulse l

Reactor Document. The average fuel temperature reached during this scenario l

can then be used to calculate an average annull gas temperature. While this scenario can be analyzed for information purposes, it is not a physically correct l

l scenario considering the presence of the coolant.

l In terms of the specific calculation, we have re analyzed this scenario and have identified an inadvertent error that yielded the 75 MWsec value and the corresponding quoted value of 3000*F average fuel temperature. The following calculation can be used to quantify a postulated " clad melt" pulse:

To raise the average fuel temperature from a nominal 100*F (at zero power) to 3365'F (zircaloy melting temperature), the adiabatic increase would be 3365 100 = 3265'F. At the hot spot where the power density is 2.92 higher than the average, the pulse energy release to yield this adiabatic rise is:

9

I NCSU PUtETAR Deceanber t$, IC]9 Response to Request for Additional leformation Revision 0 3265'F = (E)(2.92)(359 x 10r' gm)(454 gm/lb)(0.075 Btu /lb/'F)(3.413 x 10' Btu /hr/MW)(1 hr/3600 sec)

Solving for pulse energy (E) yields 70 MW5cc (versus the published value of 75 MW5cc).

It should be made perfectly clear that we do not expect clad failure at a pulse of 70 MW5cc nor do we acknowledge that an inadvertent transient can be allowed to exceed 58 MWsec (our pulse Safety Umit).

in order to remove confusion anociated with the ' clad melt" pulse limit as it relates to the 58 MWsec Safety Limit (SL). Amendment 9 of SAR Section 3.2.27 removes references to clad melt, and deals strictly with the matimum m>erage fuel i

temperature the (during an inadvertent pulse) anticipated for the fuel.

in determining the maximum average temperature rise associated with a 58 MWsee pulse at the hot spot location, the following calculation applies:

L, = 100'F + (58 MWsec)(2.92)(359 x 10' gm)(454 gm/lb)

(0.075 Btu /lb/*F)~'(3.413 x 10' Btu /hr/MW)(1hr/3600 see)

L,., = 100*F + 2707'F L,., = 2807'F The calculated value of 2807'F is less than the 3000'F used for the average fuel temperature in the calculation of annuti pressure under section 3.2.2.7. The calculation of the annull gas temperature then proceeds by assuming the numerical average between the 100*F clad and 3000'F fuel which yields 1550*F and is then approximated as 1600*F for additional conservatism. Following the calculation through, the maximum annull pressure is determined to be 372 psia.

22. QUESTIONt section 3.2.2.7, page 18 (a) Please explain better the relationship of strength of zircaloy.2 at 1600*F and at room temperature, and the conclusion stated in the last sentence. (b) Please discuss the effects of radiation exposure (damage), corrosion / erosion of the cladding, and pellet growth due to burn up on the analysis and conclusions. Discuss your fuel conditions now and for the next 20 years.

ANSWERt (a) It is invalid to consider the burst strength of the zircaloy clad at 1600*F in regards to an inadvertent pulse since the clad temperature does not approach such an extreme temperature during a pulse (due to the fact that it is submerged in the water coolant). It is more appropriate to consider the strength of zircaloy clad at temperatures slightly above bulk coolant temperatures..%n though the strength of zircaloy decreases as its temperature increases, it still provides adequate strength (and saftey margin) over the expected temperature range of the clad. For example, the 2% transverse yield stress for zircaloy at room temperatures is approximately 425 MPa, while at 1000*F the 2% transverse yield stress is 10

. NCSU PULETAR Ikcember 15,1C]9 Response to Request for Addittomal Inforination Revision 0 approximately 100 MPa. His equates into a 76% reduction in burst strength for the zircaloy for an extreme temperature of 1000*F (reference -

Zirconium in the Nuclear Industry. Franklin and Adamson editors, Sixth International Symposium, ASTM STP 824, 1984). Certainly, with no film boiling allowed in the PULSTAR following an inadvertent pulse,1000*F can represent an extreme upper reference (i.e., k temperature above the clad average temperature) that can be used for comparison of yield strength during the pulse. Yet if the measured burst pressure of 4500 psia was reduced by 76% to 1080 psia, there remains a large safety margin above the 372 psia (almost a factor of 3) calculated for the limiting pulse (in SAR section 3.2.2.7).

An additional margin of conservatism exists in this analysis from the fact that the quoted room temperature test was performed on zircaloy tubing with a 0.015 inch wall thickness. The PULSTAR actually incorporates tubing with an average wall thickness of 0.0205 inches (with a minimum of 0.0185 inches). During fabrication of the PULSTAR fuel (that we have onsite), burst testing was performed on three full length zircaloy tubes that had spacer warts brazed on and one welded end cap. These samples had l

actual wall thicknesses that ranged as follows: 0.0205" to 0.0210",0.0205 to 0.0210" and 0.02050" to 0.0215". The results were 7940, 7985, and 7965 ultimate gauge pressure (in psig) to burst the zircaloy tube samples.

l L

(b) Our present fuel loading is also our original fuel loading, but has a total power history of approximately 14,698 MWhrs (as of May 15,1989 and is i

equivalent to a burnu fuel has an average "p of approximately 1,935 MWDay/Mtonne U). Th U depletion of approximately 4.9% (Refer to SAR question # 31). Pursuant to the proposed new Technical Specification 5.1, total fuel burnup shall be limited to 20,000 MWDay/Mtonne (or 20,000 MWD /MTU). Therefore, the following comments relate to the period associated with operation with the fuel up to a burn up of 20,000 MWD /MTU (or 20 years, which ever comes first). First of all, the prototype PULSTAR at BMRC has successfully operated with fuel burnups of 20,000 MWD /MTU without any problems, meluding the disassembly and inspection of two fuel assemblies with burnups at 15,000 MWD /MTU.

One of these disassembled fuel assemblies had actually been used in the pulse program and thus represented a worst case scenario. The inspection gave satisfactory results for the fuel and authorization to increase burnup to 20,000 MWD /MTU at BMRC was granted. Similar power reactor fuel is depleted routinely to burn ups from 33,000 to 36,000 MWD /MTU with less than 0.1%' fuel pin failures. (References

• Fuel Cycle Cost Considerations of Increased Discharge Burnups", Nuclear Technology, Volume 56, January 1982 and Nuclear Fuel Management by Harvey Graves,1979). In the nuclear power industry as of 1980, more than 700,000 fuel rods had exceeded 24,000 MWday/MTU, with more than 250,000 fuel rods basing exceeded 32,000 MWday/MTU by that time. At present, utilities are extending burnups even higher, above the 40,000 MWD /MTU value. All of these power reactor fuel rods operate under much higher temperatures, i

flov/s and pressures than that observed in the NCSU PULSTAR. This extensive power reactor experience on zircaloy clad UO, operated at these l

high burnups documents the fact that the zircaloy clad can accommodate l

the radiation damage associated with a lower level of burnup to 20,000 11

~.

NCSUPUIETAR Dece2ber 15,1989 Response to Request for Addillosal luforination Revision 0 MWD /MTU burnup specific to the PULSTAR. In summary, the successful experience at the BMRC facility and the nuclear power industry in general, confirm that the fuel burnup limit of 20,000 MWD /MTU for the NCSU PULSTAR is reasonable.

In regards to the long exposure time in the core at the NCSU PULSTAR, clad erosion is not a problem since the linear flow velocities are quite i

small (approximately 1.8 feet /second). Clad corrosion is not a problem due to the low clad temperatures during operation coupled with good water chemistry. Note that simdar power reactor fuel is required to tolerate over 2000 psig pressures and 600*F conditions with various chemical additives.

In our case, the pool is at atmospheric pressure, there are no chemical additives, and the cladding is normally less than 140*F. Conditions that are conducive to corrosion of the cladding are not present in the PULSTAR.

Pellet fracture can be induced in UO, fuel by operating at high temperatures with an associated large thermal gradient across the pellet (conditions which are typical in power reactor fuel). Based on the linear power density of the PULSTAR fuel of 0.8 kw/ft, the temperature difference between the fuel centerline and pellet surface is low, on the order of 110*F at full power, and therefore, does not significantly contribute to pellet fracture.

Irradiation swelling of the fuel is induced by inert gases of fission products in the UO, fuel. With the condition of constraint (by the cladding),

irradiation swelling depends mainly on fuel burnup and the temperature of the UO, fuel pellets (reference Benjamin M. Ma, et al). As discussed carlier, the burnup for similar power reactor fuel is typically 1.5 to 2 times l

higher than proposed for the PULSTAR. As for the temperature of the l

pellets, the PULSTAR centerline temperatures are only approximately l

300 F, as compared to typical temperatures in PWRs of above 2000*F.

With the acceptable extended burnups in the power reactor industry coupled with our much lower fuel temperatures, irradiation swelling should not be a problem for the PULSTAR. Since there is some pellet growth, the gap width will actually decrease between the pellet and the inner clad surface. This effect tends to increase the gap conductance. Dilution of the Helium fill gas does however, take place with increasing fuel burnup from Krypton and Xenon, which tends to reduce the thermal conductivity of the i

fill gas. Experience with power reactor fuel has demonstrated that closing of the gap is a far more significant affect as fuel burnup increases and E

results in an overall increase in the gap conductance (reference - Ihc l

• niermal Hydraulics of a Boiling Water Nuclear Reactor. Lahey, Jr. R.T.,

and Moody, FJ.,1984)

23. QUESTION: Section 3.2.3.5.1.3-Please explain the reasons and the implications of calculations and measurements being different, bottom of page 25.

ANSWER: The value of 2.92 was calculated for a 5 x 5 fuel arrangement with only water reflection. The original core referred to as the *5 x 5 Standard Core" had this arrangement and had a measured ratio of 2.80 which is a reasonable agreement (and in any case slightly conservative of the calculated value). For the 12

NCSU PUIETAR December 18,1989 Response to Request for Additional Infonmation Revision 0 i

next two core arrangements (5 x 5 Reflected Core #1 and #3) the addition of graphite and shifting of the fuel with respect to the control rods led to smaller values of the peak to average ratio. Since the control rod / reflector arrangement is difterent than the calculation that led to the 2.92 value, they really cannot be compared. Use of the value of 2.92 as an upper limit, however, serves to allow latitude for future core arrangements where the measured value may increase slightly from the present 2.12.

24.

QUESTION: (a) Please discuss the method of control rod fabrication.

Discuss methods used to assess integrity of the control rod cladding. (b)

L Discuss the " safety feature" limiting repositioning control rods.

ANSWER:

(a) Control blade metallurgy is silver 80%, indium 15%, and cadmium 5% by weight, condition full hard, per AMF Drawing No. 89113 C-40116. Plating is nickel over tin to AMF Specification PS 2 which requires 0.001 inch minimum plating thickness. After plating, the control blade is attached to the control extension rod by five 0.250 inch diameter aluminum rivets.

Control blade plating (or cladding) integrity is assessed by periodic primary coolant delonizing resin radionuclide analysis. Gross plating wear also would be detected in the monthly radionuclide analysis of pool water.

(b) The control blade moves within an aluminum shroud. This shroud has two

? ns on its base that insert into core grid plate holes. One circulcr dowel i

dowel Ls used to position the shroud on the core grid plate, while a second semi-circular dowel prevents the shroud from rotating. Both of these dowel pins have diameters larger than the narrow width of the control rod shroud. When the shroud is positioned correctly, the dowel pins become I

l fully inserted into the grid plate such that the area immediately around the outline dimensions of the control rod shroud is clear. The four fuel l

assemblies that are positioned adjacent to the shroud actual rest on top of these dowel pins. Therefore, to remove the control shroud (and blade),

one must first unload the four assemblies that surround the control rod shroud.

25.

QUESTION: Section 3.2.3.5.3 (all sections) - (a) Please discuss differences between NCSU and BMRC, and between predictions and actual measurements.

(b) 3.2.3.5.3.5 - Please explain whether these fuel assembly worths are for fresh fuel, and explain the changes.

ANSWER:

(a) The measured moderator coefficient temperature for BMRC was higher than the NCSU PULSTAR due to its smaller core size (5 x 4 array versus 5 x 5 array) and different fuel / water ratio (as a result of 6% enrichment versus 4%). As additional graphite as been added to the PULSTAR core, the coefficient has become more negative due to reduced leakage.

The predicted Doppler coefficient for the NCSU PULSTAR was 2.85 pcm/'F while the measured value (for the 5 x 5 Standard Core) was -1.4 pcm/'F. The higher BMRC value is the result of a " harder" spectrum 1

13

NCSU PUtRTAR Decesiber 18, IC 9 Response to Regnest for Additional leforatation Revisloa 0 j

1 associated with the higher "U enrichment.

De difference between the predicted and actual total predicted loss for steady state operation follows the differences associated with the fuel and l

moderator temperature coefficients just discussed.

Since the void coefficient of the NCSU PULSTAR was estimated from Buffalo measured data, the measured value differs by a small amount.

His difference is not sur) rising since the measured value of the moderator temperature coefficients ciffer between our facilities.

The original 5 x 5 Standard Core had a predicted peak ecuilibrium xenon of 1060 pcm. While this core never operated for longer taan 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> at full power, the 5 x 5 Reflected Core #1 was operated to reach equilibrium xenon and the measured value was approximately 850 pcm. This agreement is reasonable, (b) The quoted data is based on the actual burnup of the fuel assembly that existed at the time of the Start up test for the core. Since the 5 x 5 Standard Core and 5 x 5 Reflected Core #1 arrangements do not have enough useful excess reactivity to operate, only the 5 x 5 Reflected Core

1. 3 applies toward future operation. The location of the highest worth assembly changed from the 5 x 5 Standard Core / 5 x 5 Reflected Core

1. 1 position (grid location 3 C) to the present location (grid location 4 D) due to the movement of the fuel assemblies on the west grid plate row to the most east grid plate row associated with the installation of the 5 x 5 Reflected Core #3. The decreasing value of the highest worth fuel assembly from the original core to present is due to the contribution of the graphite reflectors.

26.

QUESTION: Section 3.2.4.1.1, page 33, Steady State - (a) Are criteria (2) and (3) also for forced convective flow? Discuss. (b) An additional design basis should be no fuel melt in case of LOCA or any credible transient, or other accident. Please address this.

ANSWER:

(a)

Appendix 3B, part III.C.1, page III 12, of the SAR details the analysis i

results of the steady state heat trarufer behavior that is required to meet the design criteria of SAR Section 3.2.4.1.1. In response to your specific question, criteria (2) and (3) do apply to forced convective flow.

(b) Certainly, no fuel melting is a fundamental design basis for accident scenarios such as a LOCA or inadvertent pulse. As required by SAR Section 3.2.4.2.1, " Transient Heat Transfer Analysis - Design Criteria", no fuel melting in the case of pulsing is allowed. The inaMrtent pulse safety analysis as detailed in SAR Section 13.2.2 does not exceed the 58 MWsec Safety Limit on the fuel for which the design basis of SAR Section 3.2.4.2.1 applies. In the case of a LOCA, the design basis includes no fuel melting and no clad melting. Based on the analyses in SAR Section in 13.2.1.4, both of these design criteria are achieved, since the surface fuel temperature is only expected to reach 765'F. Refer to our response to SAR question # 61.

14

NCSU PULSTAR Decesiber 15,1989 Response to Request for Additional leforniation Redsloa 0

27. QUESTION: Section 3.2.4.1.3.1, pages 35 and 36 - (a) Fuel loading and pin spacing; discuss whether these variations were found to exist in the as built reactor, and implications. (b) Instrument Error: is this d.7 percent some sort of limit, root mean square, or what? How is it verified?

ANSWER:

(a) PULSTAR Fuel O.C. Document provided by Canadian Westinghouse Company Limited indicated that the variation in fuel load among the delivered fuel pins was from.572 to.581 grams UO, (or a variation of 1%

to +1.15% from the average). This is well within the 2 2% allowed by SAR Section 3.2.4.1.3.1.

PUISTAR Fuel Q.C. Document indicated that all of the zircaloy tubing fabricated and used, fell within the allowed dimensional tolerance of i 0.0015 (which is derived from the 0.474" + 0.00" -0.03" outside dimension requirement) In addition, fabrication records indicate that the

'i fuel assembly box inside dimensions also fell within the allowed dimensional tolerance of 0.005".

l (b) The 7% is a design upper limit. The actual accuracies of the full power neutron detecting channels (as detailed in the instrument's electrome design specifications) are:

Linear Power Channel -

2% up to 1 MW detector current and 2596 i

over 1 MW detector current for the amplifier and an additional 0.25% for the recorder output.

l Safety Power Channel -

2% up to 1 MW detector current and 5?t l

over 1 MW detector current for the amplifier and an additional 0.25% for the recorder output.

I.og N Intermediate Range - 4% for the amplifier electronics (originally was 4% but system electronics were replaced to remove vacuum tube technology based components) and an additional 0.25 %

for the recorder output.

It is appropriate however to note that the ultimate power measuremene that is used to calibrate the neutron detectors is based on a heat balance.

In the heat balance, the instrument errors involve a 0.1'F tolerance with respect to the absolute temperature and 0.1'F maximum difference tolerance between the inlet and outlet RTDs (when at isothermal conditions) using a NBS (now referred to as NIST) type thermometer measurement compared to the Temperature Recorder output. This allowed tolerance yields a 0.7% maximum error at full power assuming the delta temp across the core is 13.8'F

28. QUESTION: Section 3.2.4.1.3.2.. page 38 Please discuss your reasoning that limiting "N release is the fundame ital reason for no bulk boiling in the reactor e

NCSU PULSTAR December 15, 1989 Response to Request for Additional Infonnation Re M om 0 core.

ANSWER: Avoiding DNB in either forced convection or natural convection cooling is the paramount design criteria. However, stable bulk boiling can occur without the onset of DNB, and this is the actual heat transfer mechanism in Boiling Water Reactors (BWRs), in our case, it was decided to further limit convective heat transfer to no bulk boiling. In this way, "N rise to the surface (aided by quick bubble rising) can be avoided. In summary, we have chosen a more limiting design criteria than by just considering DNB.

5

29. QUESTION: section 3.2.4.1.3.2, page 39 and 40 Please provide a quantitative discussion of the " chimney effect."

ANSWER: A chimney is an unheated extension of the fuel assembly housing that contains no dividers or internal parts. Test results that are used in support of the NCSU PULSTAR do not take credit for the increased performance that is observed using a chimney. The addition of a chimney of height H. increases the driving pressure by the quantity:

(p.

r.)H,.(g/g,)

Where r, is assembly inlet density and r, is the assembly outlet density, and the ratio g/g, is numerically 1.0 in standard gravity. The increased driving pressure will result in greater natural circulation flow for the fuel assembly than that observed without a chimney.

30. QUESTION: Section 3.2.4.2.3, page 45 - (a) Was onset of DNB observed at 58 MW see? (b) NRC would take a dim view of planning to operate with DNB.

Discuss probably DNB ratio.

ANSWER:

(a) The prototype PULSTAR test document under reference 31 assumed that

" pin bowing and the film boiling threshold are coincident" and therefore provides a method that correlates the onset of measurable pin damage (i.e.

bowing). For the NCSU PULSTAR at 400 gpm flow (80% of nominal), it is predicted that film boiling shall not occur below an energy density of 470 l

wattsec/gm which is equivalent to 58 MWsec during a pulse.

l (b) The minimum DNBR during routine operation is 2.0 (either in natural circulation or forced Dow) as stated in SAR section 3.2.4.1. The Safety Limit as defined by ANSI 15.1 1982 is *.. limit on process variables which are found necessary to reasonably protect the integrity of the principle physical barriers which guard against the uncontrolled release of radioactivity... and...the principle physical barrier is often the fuel cladding.

Since pulsing is to be removed from our operating license your 58 MWsec concern only applies to a startup accident (either by rod withdrawal or experiment reactivity). At the equivalent energy density of 470 wattsec/gm, only slight bowing was detected on the test fuel pins (reference 31).

Furthermore, test pins were pulsed up to energy densities as high as 872 wattsec/gm. Therefore, in the case of the NCSU PULSTAR, the 58 MW5cc represents a reasonable upper limit of energy release during the 16 l

m NCSU PU1KTAR December 15,1989 Response to Request for Additional Information Revision 0 l

pulse to avoid damage to the physical barrier (to avoid an uncontrolled release of radioactivity).

31. QUESTION: Table 3.2 - (a) " Savings" for a H,0 reflector is given. Please compare with a graphite reflector. (b) What fraction of the quoted temperature

]

coefficient of -3.9 pcm/'F is prompt? (c) How is the void coefficient characterized? Is the value listed a core average, or is it for a specific core location? Please discuss. (d) K., - core (1) Does " rods out" mean all rods out including the Pulse Rod? (2) What is the maximum variation in actual U 235 burnup among individual fuel assemblies?

ANSWER:

(a) While the graphite reflector " savings" is not available, the measured worth of the graphite was determined during startup test for the 5 x 5 Reflected Core # 3. Measurements indicated that the graphite reflectors in grid positions 1A to SA represented a gain of 960 pcm over a water reflector.

The graphite reflectors in grid positions 6A to 6E represent a gain of 1020 pcm over a water reflector. However, the net gain in excess reactivity that was measured in going from the 5 x 5 Reflected Core # 1 (which had 5 l

graphite reflectore to 5 x 5 Reflected Core # 3 (which had ten graphite reflectors) wa=

pcm. Refer to SAR Figures 3-8A and 3 8C for core t

l configuratior (b) We interpro to mean "what contribution of the -3.9 pcm/'F is caused by a rise b

.iel temperature (since it rises accordingly with bulk pool the contribution is a -1.6 pcm/*F from Doppler. Note that tempers L

the mode temperature coefficient is really an isothermal moderator E

coefficient, where the pool and fuel temperature are raised and the-observed reactivity change is recorded.

(c)

The void coefficient for a particular core is measured in a coolant channel that has the aversge power density (or neutron flux). For example, with the 5 x 5 Reflected Core #3, the average flux existed in grid position 4B L

l and coolant channel Al'of that fuel assembly. By comparison of the critical rod height between a water filled " wand" and a air-filled wand (approximately 5 cm' volume) inserted in the coolant channel at low power, the void coefficient was calculated.

(d) (1) " Rods Out" does indeed refer to all the scram capable rods (Safety #

1, Safety # 2 and Regulating Rod) and the Pulse Rod (or " slim" rod) being fully withdrawn from the core.

(2) As of May 15,1989 with a total of 14698 MWhrs historgon the original 25 fuel assemblies, the average burnup is 4.9%

U depletion, with a maximum of 6.4% and a minimum of 3.4%.

32.

QUESTION: Table 3 Please discuss the reasons for differences between predicted and measured void = effects. What implications are applicable to other reactivity calculations performed with the same methods and computer programs?

ANSWER: Because of fuel / control rod / reflector changes, the 5 x 5 Reflected 17

NCSU PUESTARL December 18, 1989 l

Response to Request for Additional Information Revision 0 Cores #1 and #3 cannot be compared to the calculated values for the 5 x 5 Standard Core. In regard to the 5 x 5 Standard Core, the predicted values for the (3) 6" beam tubes are reasonable, i.e. 200 pcm versus an average of 193 pcm. For the 8" beam tube, 6" thru tube, and the 12" beam tube, the calculation assumed that the beam tubes extended up to the fuel surface, whereas in reality in each of these cases they do not. For example, on the 5 x 5 Standard Core, the 12" beam tube has a water gap approximately 3" thick between the closest row of fuel and the beam tube endplate. Therefore, the predicted value of 1000 pcm overestimated the actual 20 pcm worth of the tube.

Similarly, there exists a gap between the end of the 8" beam tube and the fuel surface, and contains the startup source holder and the pneumatic tube.

Therefore, its measured worth of 70 pcm-was lower than the predicted 200 pcm.

The original core nuclear calculations would not have been affected by the assumptions in the beam tube worth calculations.

33. QUESTION: Section 4.2.3 - Discuss whether methods to prevent or avoid corrosion were fully successful. Discuss implications for the next twenty years.

ANSWER: Yes, we do feel that our corrosion control program has been successful, even in light of our recent liner leak, Samples of the pool liner material have been suspended in the pool water since startup and serve to document corrosion of the liner. Measurement taken at six months intervals t.

L during the first 5 years of operation, and subsequently on an annual basis have yielded no detectable weight loss on these " coupons". Monthly analysis of the Primary Coolant is designed to identify and quantify corrosion products. This process consists of Neutron Activation Analysis (NAA) of 5 ml of pool water and is compared to a 5 ml standard solution of Nacl. This analysis has consistently provided results considerable below the 1 part per million (ppm) limit.

I L

The isolated leak site in the liner (identified in February of 1988) was likely the result of a fabrication flaw, whereby some foreign material was embedded in the liner material. It is our understanding that the pool storage pits were fabricated by rolling flat plate into the form of a cylinder and did experience repairs (grinding and welding) prio: :o installation at the facility (after failing leak testing). During the testing prior to restart of the PULSTAR Reactor after the leak initiated, general ultrasonic testing of the liner was performed and indicated that the original full thickness was present. It was our conclusion that the leak site was an isolated event and not indicative of liner performance in general.

In regards to the known leak site, preparations to permanently seal the defect is in progress in cooperation with National Nuclear (NNC), Limited. As of this time, a remote operated device has been constructed to position an epoxy resin patch over the leak site area with mockup testing to occur in the near future.

The radiation damage characteristics, " pull-off strength", and chemical leachibility of this proprietary repair sealant are well documented by NNC and therefore was deemed to be the ideal permanent repair of the minor leak site.

In terms of the future, maintaining our existing water chemistry with our L

successful results, in conjunction with the surveillance of Primary Coolant water, L

provides the basis for the facility to proceed with an additional twenty years of I

operation.

f L

NCSU PULSTAR Decernber 15,1989 Response to Request for Additional Infonnation Revision 0

34. QUESTION: Section 4.2.1 Under what conditions is the water pressure higher in the primary side or secondary side of the heat exchanger? Discuss in relation to possible leakage across the barrier, and control of radioactivity. If leakage to the secondary side is possible, discuss activity and isotopes that could be released to the environment.

ANSWER: With the Secondt.ry Pump secured and the Primary Pump either secured or operating, the pool water elevation is higher than the cooling tower a

basin level and therefore a leak in the heat exchanger would flow from Primary to Secondary. With both the Secondary Pump and the Primary Pump operating, pump discharge pressures will cause leakage to go from Secondary to Primary.

If a leak developed under these operating conditions, Secondary coolant additive would also migrate to the Primary,.which would serve as a flag to the reactor operator resulting from reduced Primary coolant resistivity and increased Primary coolant radioactivity (which would appear at the demineralizer resin bed).

L Monthly analysis of Secondary coolant are made to ensure that no Primary radioactivity is present in order to address the times when the Secondary System L

is secured. Monthly Primary coolant analysis is made to detect not only fuel problems, but also Secondary System chemical additives that could have leaked p

m. To operate at power, the Secondary System is required to remove heat in l

order to maintain a constant pool temperature. The Primary coolant activity is highest while the reactor is operating at power and decreases after shutdown by way of the purification system. Therefore, the leakage from Primary to

~ Secondary during the times when radioactivity is the highest in combination with coolant analysis to detect trace amounts of either radioactivity in the Secondary coolant or Secondary coolant chemical additives in the Primary coolant, address the control of radioactivity. Sodium is by far the main contributor to primary coolant radioactivity, with small contributions from manganese, antimony, and zinc.

l

35. QUESTION: Fig. 4.1D - What method is used to prevent primary coolant from l

entering the water supply system of the engineering building under any pressure conditions or valve leakages?

l ANSWER: Water additions to the Primary System are made in a batch" mode l

where the Service Water System is connected to the Primary by way of a

• Quick-Disconnect Fitting" (refer to SAR Figure 4-1D). The Service Water System is brought on-line and the raw water circulated through the filters /demineralizer to achieve the required water quality. This makeup water is then injected into the Primary Pump suction (refer to Figure SAR Figure 4-L 1A) by way of the quick disconnect fitting. Therefore, the potentially radioactive Primary System coolant is not routinely connected to clean systems such as the i

Service Water System. The typical raw water pressure on campus available supplying the Seuite Water System is 75 psig. Because of this pressure difference water will always flow to the Primary System.

Furthermore the Service Water System connection to raw water contains two check valves in series to prevent reverse flow into the raw water (again, refer to SAR Figure 4-1D). The combination of check valves, system pressure difference, and routinely not being connected ensure that Primary coolant cannot migrate to raw water.

L 19

NCSU PUIS1'AR December 15,1989 Response to Request for Additional inforenation Payisloa 0 l

36. QUESTION: Section 4.23 - Provide a quantitative discussion of the statement in the last sentence of Section 4.23 about fuel assembly leakage.

ANSWER: A major breach in the cladding of a fuel pin would quickly appear on both the N Power Measuring Channel and the gaseous effluent monitors.

In terms of the N Channel, the increased radiation emanating from the outlet piping would increase the indicated detector current which normally is directly proportional to total reactor power. The Reactor Operator would see this as a mismatch between the normally corresponding unear and Safety Channels with the N Channel. The gaseous release associated with a major breach in the fuel cladding would cause the radiation alarms to sound and automatically place the Reactor Bay into the confinement mode.

A fuel assembly that develops a small leak would likely be detected within the first twenty four hours by the primary demineralizer radiation monitor. The size of the leak and the age of the resin would dictate when the primary demineralizer radiation level would increase by a suspect amount of 20 mR/hr over normal values. The scenario would be that the leaking fission products are pumped through the delay tank and into the purification system and produce a gaseous release from the pool surface that does not trigger a radiation monitor ALERT or ALARM. As the fission products are collected in the 8 cubic feet of mixed bed resins, the primary demineralizer radiation monitor dose rate would increase accordingly and be noticed by Health Physics staff and/or the reactor operator. This scenario appears most probable for a small leak, especially in light of the details from a recent report of a leaking fuel assembly at the University of Michigan Ford Reactor.

There are several other avenues a leaking fuel assembly could be detected. For instance, the Continuous Air Monitor (CAM) with a charcoal and particulate filter is positioned down wind (e.g., associated with the HVAC air flow) of the pool parapet to allow sampling of air over the pool surface. Fission products migrating to the surface that become airborne would be detected by this system.

The monthly water analysis of 250 ml of Primary water includes gamma spectroscopy to detect fission product isotopes. And finally, analysis of 5 ml of spent resin following resin change (typically every six months) also provide a method for detection of fission product activity.

The particular system or technique that would initially detect a leaking fuel pin would be dependant on the size of the clad failure, the age of the resin, and the history of the reactor. The combination of the surveillance recent operating / resin and the console instrumentation (i.e., gaseous radiation analysis of water channels, N Channel, etc.) provide a comprehensive methodology for detecting fuel pin failure.

37. QUESTION: Section 5.2.1, 5.2.2 - Please provide a detailed quantitative discussion of the levels of radiation that trigger events, and their bases. Please state and justify any assumptions used.

ANSWER: The attached Table 7-1 of the PULSTAR Operations Manual (POM) detail the various radiological channels and their actual setpoints. The ALERT and ALARM setpoints on the stack gaseous monitor was chosen to be

, 20

- ~. -

NCSU PULSTAR December 15,1989 Response to Request for Additional Information Revision 0 20% and 80% of 10 MPC for *Ar, respectively, that could occur at the D.H.

i Hill Library upper floors. The ALERT and ALARM setpoints for the Stack Particulate are based on 20% and 80% of 10 MPC for?re, also at the library

)

location. Using a Gaussian Plume model and the predominant meteorological pattern, D.H. Hill was postulated to be the limiting site in proximity to the PULSTAR Reactor facility (refer to Amendment 9 of SAR Section 13.2.1.5),

with a minimum dispersion factor (at Stability Class "F' and a wind spped of 1 m/sec) of 113 between the top of the stack and the library itself. The ALERT and ALARM settings for the area monitors were chosen based on occupational and non-occupational areas. The Control Room channel ALERT is at 2.5 mR/hr, the lower limit of a radiation area. All other monitors are non-occupational and have an ALERT setting of 10 mR/hr and an ALARM setting of 100 mR/hr (the lower limit of a high radiation area). The span between the ALERT and ALARM setting is provided for remedial actions as appropriate.

38. QUESTION: Section 5.2.2 - What indication does the operator have that dampers are fully closed? Will the operator know if the damper sticks while closing?

ANSWER: The normal ventilation isolation dampers (inlet, exhaust, Control Room), must fully close before their status light will energize on the Radiation l

Alarm Panel and therefore is a "not. closed" status light. The confinement fan isolation dampers must fully open before its status light will energize and therefore is a "not open" indication. If the dampers do not reach their destined position due to sticking when confinement is initiated, the respective status light will not energize. This information is readily available to the reactor operator L

to verify _ correct damper position for confinement.

39. QUESTION: Section 5.2.2 - Can the dampers close automatically without any power assistance?

ANSWER: As stated in SAR section 5.2.2, the normal ventilation isolation dampers automatically close upon loss of either commercial power or compressed air supply. The confinement isolation dampers are electrically powered and will fail to the closed position upon loss of power. Note that the L

confinement isolation dampers and the confinement fan motors are supplied from the same source and therefore can be powered by the Auxiliary Generator upon loss of commercial power and therefore can be placed in operation.

40. QUESTION: Section 6.1 - Is there an indication to the operator that the flapper is fully open? Is it possible to operate the reactor above 150 kW with the flapper not fully closed?

ANSWER: The indication available to the reactor operator is that the flapper is not closed and is determined by a microswitch on the reactor bridge coupled to an extension leading to the flapper. The "not closed" measurement is used versus " fully open" in a conservative approach to initiating a reactor SCRAM when the potential for inadequate flow through the core exists. Because of the Flow-Flapper enables that exist on the Log N Channel (Intermediate Power Range) and the Safety Channel (redundant Power Range), the reactor will SCRAM if the Flapper is not closed at or above 150 kw (Safety System Setting).

l 21

NCSU PUIETAR -

I)ecember 15,1989 Response to Request for Additional Information Revision 0 The 250 kW value is the limiting Safety System Setting for these enables.

41. QUESTION: Section 7.1.1 - Please provide the ranges and uses of the period meters in your control console.

l ANSWER: The scale of the Startup Channel Rate Meter covers the period range of -100 seconds through infinity (=) to a full scale of +10 seconds. The instrument has two built in test inputs at +10 seconds and +30 seconds. The ability of the instrument to accurately display these two test inputs is routinely verified. The I.og N Channel Startup Rate Meter has a range from -0.5 dpm (decades per minute) to +3.0 dpm.

42. QUESTION: Section 7.1.1 - What is the function of the "high voltage supply,"

and what is the purpose of the interlock?

ANSWER: The internal "High Voltage Supply" could have been used to power the Start-up Channel fission chamber detector but our facility elected during construction to provide a local high voltage power supply at the Reactor Bridge.

This internal supply therefore is not used. However, the Startup Channel readout chassis was designed to verify that this supply is available (based on the assumption that it would be used) and therefore interlocked it into its operation.

We have elected to leave this interlock as is, since it routinely does not affect its operation. If it were to fail, we would inspect and repair the unit to verify no other components in the chassis were damaged as a result of the failure.

43. QUESTION: Section 7.4.1. - What are the actions that are permitted and/or inhibited by the various positions of the Gang Drive Switch?

ANSWER: The three control rods, Safety #1, Safety #2 and Regulating Rod must first be selected "On Gang" by depressing a button at the individual rod

-drive switch. One, two or three of these rods as chosen by the reactor operator can be operated by the Gang Rod Drive Switch. Whether the individual control rod is selected on the Gang Rod Drive Switch or not, its individual rod drive switch is fully functional within the' limitations of having magnet power, no Startup Channel Inhibit (refer to SAR Section 7.4.3), Ganged Insert Switch to the OUT position and no Reverse / Scrams present. Note that there exists a control referred to as the Ganged Insert Switch (which is used a convenience to drive the rods in on a reactor shutdown and has no capability to withdraw rods) and the Gang Rod Drive Switch which is the subject of this question. Further statements will now apply assuming the rods have been selected on gang and are operated by the Gang Rod Drive Switch. Rod movement into the core is never inhibited and always fully functional. Rod movement out of the core by the gang switch can be stopped by: a Startup Channel Inhibit, no magnet power, a Reverse and or SCRAM signal present, Ganged Insert Switch not in the OUT position. Gang Rod Drive Switch movement is not allowed while the reactor is under Automatic Control. Attempts to position the Gang Switch while in Automatic, will result in the Automatic Channel disengaging back to manual control.

44. QUESTION: Section 7.4.2 Page 19, Paragraph 1 - How is "I.ow Shutdown Margin" monitored and interpreted by the instruments?

22

1

- NCSU PUIETAR Decenter 15, 1989 Response to Request for Additional Information Revklom 0 l

ANSWER: A k,w shutdown margin is assumed by the reactor instrumentation if the LOG N Operative button has been depressed (To transfer from Startup L

Channel to the LOG N Channel) while any of the three control rods are below a preset rod height. The assumption by the instrumentation is that the reactor would have to be critical /supercritical to require the need for transition into the IDG N Channel. An additional " benchmark" is used by the operator to ensure adequate shutdown margin during a reactor startup. As detailed in PULSTAR t

Data Summary Volume II, the rod height of 13.1" corresponds to twice the required shutdown margin (800 pcm) when certain assumptions are met. Refer to PULSTAR Operations Manual Section 3 for details.

45.

QUESTION: Table 7 Please be certain that all of the protective channel settings are appropriately included in your technical specifications. (1) Items 5 and 6: text seems to say 1.2 MW, not 1.3 MW. (2) Are items 10 and 11 correlated? (3) Items 12 and 13: text seems to say 116*F.

ANSWER:

(1) The value of 1.2 MW is the actual Safety System Setting while 1.3 MW is the Limiting Safety System Setting. SAR Table 7 3 has been revised to clarify the Safety System Setting versus the Limiting Safety System Settings.

(2) No, they are diverse methods of ensuring adequate pool level, but are not correlated. The 100 mR/hr is based on radiation personnel protection (such as from removing irradiated samples or from low water level) and is not equivalent to 14 feet,2 inch level.

(3) Again,116 F is the Safety System Setting, while the value of 117 F is the Limiting Safety System Setting.

46.

QUESTION: Chapter 8 - What area or effluent monitors are on auxiliary electric power? Discuss reasons.

ANSWER: All area and effluent monitors displayed on the Radiation Monitoring Rack (on the console) can be powered by the Auxiliary Generator once the " Console Distribution" load is transferred to the generator. The three Area monitors (Control Room, Over-the Pool, and West Wall) have battery back-up capability that will last approximately 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> and is available immediately upon loss of commercial power. The VAMP located on the reactor bridge has a battery back-up, but cannot be powered by the generator. The Constant Air Monitor (CAM) located on the Reactor Bay also cannot be powered by the generator. The availability of all the Radiation Monitoring Rack readouts ensures that adequate radiological information is available concerning area dose rates and gaseous effluents upon loss of commercial powec.

47.

QUESTION: Section 9.1.4 - How is the water level in the liquid waste holding tanks determined?

ANSWER: The level in the waste tanks is determined by a float device with linkage coupled to a potentiometer. The resistance value of the potentiometer is used to transmit a signal to a remote readout in Room B103. The readout in

. 23

4

-t NCSU PULSTAR December 15,1989 Response to Request for Additional Information Revision 0 B103 is scaled from 0 to 100%

48. QUESTION: Please discuss and describe your fire protection program and provisions, including training and maintenance.

ANSWER: The Burlington Engineering Labs (BEL) has a Honeywell based fire protection system that is under the authority of the university's Physical Plant organization. This group performs repairs / maintenance on the system and tests the fire detectors. In the event a fire signal is generated (either by a pull-station or a detector initiation), the Raleigh Fire Department is automatically summoned and is escorted through the NCSU campus by the Public Safety organization. Only the Physical Plant organization can reset the system after actuation. Fire extinguishers are maintained and tested by the university's Life Safety organization. The training of Fire Department personnel on the PULSTAR Reactor Facility is handled via the Emergency Plan (and Procedures) on an biennial basis. Training of PULSTAR Reactor operator response to a fire condition is handled via the Requalification Program, using the PULSTAR Operations Manual detailed fire response procedure as the training guide.

l 49.

QUESTION: Section 10.1.2 - Please provide a more quantitative discussion of L

releases and dilution of liquid wastes before and after release.

1 1

ANSWER: Health Physics Procedures HP 20-2, " Release of Radioactive Water j

into the Sanitary Sewer System", and HP 20-3, " Sampling Waste Water Tanks" are attached to provide the requested quantitative discussion.

50. QUESTION: Section 10.1.3 - Additional parts of 10 CFR apply to waste disposal other than Part 20. Discuss your reasons for only citing 10 CFR Part 20.

ANSWER: All low level solid waste generated by the PULSTAR Facility is transferred to the campus state license for ultimate disposition. The campus Radiation Protection Office is the university's' North Carolina State license holder and serves to coordinate disposition of all low level solid waste from the various campus sources (the PULSTAR Reactor being one of them). Since North Carolina is an " agreement state",10 CFR 61 is applicable for low level waste disposal, and 10 CFR 70 is the pertinent regulation for shipping activities associated with waste disposal.

l-51.

QUESTION: Section 10.1.4 - Chapter 5, this section of Chapter 10, and your environmental report does not provide sufficient treatment of Ar and "N.

l Please provide a detailed, quantitative analysis.

ANSWER: Ar produced during normal operation is purged from the PULSTAR Reactor Bay by the ventilation system and is exhausted out the 100 foot stack. Over the last ten years, an average of 6.08 x 10* cl/cm' of Ar was exhausted from the Reactor Bay and diluted by an approximate factor of 2 before leaving the stack (based on the dilution provided by the BEL Old Wing ventilation system). Therefore, the average release for the last ten years was 3.04 x 10* ci/cm'. Even though the stack top is inaccessible by the public, the unrestricted release limit for Ar was still met. The Ar is produced by: air 24

NCSU PUt3 TAR Deceinber 15, 1989 Response to Request for Additional Infonnation Revision 0 dissolved in pool water (9.4%), Beam Tubes and Thermal Column (7.1%), and Pneumatic system (83.5%). To minimize Ar production, the pneumatic system is only operated when required for sample transfer, and otherwise purged with nitrogen to remove the air.

The dose rate of N traveling through the pool. outlet piping has been measured with the reactor at full power and Primary System at rated flow, and was determined to be 8 R/hr (on contact with the pipe). The measurement was taken in the valve pit adjacent to the biological shield, but is normally not accessible due to the pit shield plugs. The primary piping leaving the Reactor Building is a minimum of 8 feet underground with a residence time under ground of greater than 1 minute. This allows for a minimum of 9 halflives decay of N before the primary piping re-enters the building. Evidence of N as the piping emerges into the Mechanical Equipment Room (MER) is not

-detectable. During forced flow conditions at full power, the dose rate measured above the pool averages 1 mR/hr and originates primarily from the core with a minor contribution from Ar.

N produced around the periphery of the core

'(i.e., not drawn into the core and subsequently, the outlet piping) migrates about the core and slowly rises, decaying as it goes and does not contribute to the pool surface dose rate. In natural convection flow at 10% power the major pool surface dose rate contribution is from N rising with the convection water currents and totals approximately 3 mR/hr.

Amendment 9 of SAR Section 10 now includes this and additional detailed additional analysis concerning both Ar and N, 52.

QUESTION: Section 10.2.1 - Please provide a quantitative discussion of the radiological conditions associated with operating at full power with 14 ft. 2 in. of water above the core.

ANSWER: With the reactor at 95% power in forced convection flow, measurements indicate that the gamma dose rate at the 14 feet 2 inch level is approximately 110 mR/hr. Based on this measurement it is predicted that the dose rate at the pool surface would be approximately 50 mR/hr (with the pool level at 14 feet 2 inches). With the reactor at 10 % power in natural convection flow, measurements indicate that the gamma dose rate is approximately 250 mR/hr at the 14 feet 2 inch level. Based on this measurement it is predicted that the dose rate at the pool surface would be approximately 114 mRfhr (with the pool level at 14 feet 2 inches). These measurements provide conservatively high values, since back scattered radiation from above the 14 feet 2 inch level of the measurement was also detected.

Since our Safety System Setting (SSS) is set at 17 feet, radiological conditions associated N and Ar are manageable. Note that we have our ALERT setpoint on the Over the Pool ion chamber set at 10 mR/hr (and the VAMP set at 2.5 mR/hr), and we have never had either of these alarms to energize from routine operation during the facility lifetime, either in forced or natural convection flow.

53.

QUESTFON: Chapter 10 - Describe in more detail the organizational structure of the J ersanne protection organization, and discuss qualifications, training, origin C prov.mres, relationship to reactor manager, etc.

25

l h.i L

NCSU PUUiTAR

. Deceinber 15, 1989 F

Response to Request for Additional Inforination Revision 0 i

i-1 ANSWER: The requested information is detailed in the attached document,

" Handbook for Protection Against Ionizing Radiation," particularly, section 1.

This document details the campus organizations involved with radiation protection and their interactions. Note that the " Nuclear Operations l

Administrator" title used in this document has been replaced with " Associate Director"

54. QUESTION: Chapter 10 - Please provide a summary of the radiation exposure l:

histories at the NCSU Reactor Facilities, in a format outlined in 10 CFR 20.407(b).

ANSWER: The requested radiation exposure histories at NCSU in 10 CFR 20.407 (b) format is attached.-

55. QUESTION: Section 11.1 and its subparts - Please provide a document that establishes an ALARA philosophy and policy in the University to assure compliance by all staff and users of the reactor facilities.

ANSWER: The requested ALARA document is attached for your review.

56.

QUESTION: Section 11.2.1 and 11.2.4 - Discuss the need for reactor operator training to meet the requirements of 10 CFR Part 55.

ANSWER: Our facility recognizes the fact that 10 CFR 55 is the law

- concerning reactor operator training requirements (including requalification) and we therefore comply accordingly. Both Special Procedure 2.3, " Reactor Operator Qualification", and Special Procedure 2.6, "PULSTAR Operator Requalification Program", were developed and will be continually revised to meet the requirements of 10 CFR 55.

57.

QUESTION: Section 13.2.1.1 - For your 'Ar releases, please provide a detaileo d

quantitative analysis of potential exposures and doses to personnel in unrestricted areas, including the most-exposed offsite permanent residents and any affected l

campus buildings. Please include the dispersion / diffusion factors you have used in the analysis. Also address potential exposures in the reactor room. Clearly state and justify any assumptions used.

l ANSWER: Amendment 9 of SAR Section 10.3.2 now details the radiological consequences to offsite locatiens of the routine Ar releases associated with the PULSTAR Reactor operation.

The calculated concentration of Ar in the Reactor Bay during normal power operations is 1 x 10* Ci/cm'. This data is derived by using an average release of Ar at the stack exit of 3.04 x 10* Ci/cm' (or an average of 6.783 Curies / year). The concentration of d'Ar is lower in the Bay than at the stack top because the pneumatic sample transfer system generates approximately 83.5%

of the Ar and dumps directly into the HVAC system, rather than the Bay. The dose associated with this Ar concentration in the Bay is calculated to be 0.0377 mrad /hr gamma and 0.0133 mrad /hr beta. Actual exposures to reactor staff are L

further reduced by the fact that the Bay is not continuously occupied.

t

• 26

R

.NCSU PUtSTAR December 15,1989 Response to Request for Additional information Revision 0

58. QUESTION: Section 13.2.1.2 - What is the affect of a failure of the Dapper to open in the loss of flow accident?

ANSWER: As stated in section IV. A. 2. of Appendix 3B of the SAR, the analysis assumes that the core flow stops instantaneously at the time of the transient. The film coefficient for heat transfer was also assumed to drop to a conservatively low value representative of pool beat transfer conditions.

Furthermore, as stated in this section, it was assumed that during the transient the film coefficient remained at this low value and the Dow was not affected by natural convection effects which would occur once the Dapper is open.

Therefore, the failure of the flapper to open in the loss of a flow accident is encompassed by the safety analysis as detailed in SAR Appendix 3C, section IV.

A. 2.

59. QUESTION: Section 13.2.1.3 - Please provide a quantitative analysis of waterlogging, because inadvertent transients are credible events.

ANSWER: A postulated waterlogging accident in the PULSTAR Reactor would involve the unrelated events of a fuel pin cladding breach and an accidental transient (i.e., pulse). SAR Section 13.2.1.5 has analyzed the case whereby an accidental release occurs from three separate fuel pins by mechanical damage and demonstrates that the consequences of this postulated failure are manageable. In the case of waterlogging, fission products would likely be detected in the Primary Coolant after a breach of cladding that was large enough to allow water to enter. If a fuel cladding breach was suspected, the PULSTAR staff would work expediently to identify and remove the fuel pin from service to avoid further contamination of the pool surfaces (that would occur for a very small cladding leak). Therefore, the three-fuel pin failure analysis of SAR Section 13.2.1.5 provides a reasonable upper limit of a release that could occur with a postulated waterlogging accident.

60. QUESTION: Section 13.2.1.4 - Please describe and discuss in more detail the penetrations in the pool liner, and.the primary coolant system by which pool drainage could occur accidentally. Include an analysis of a primary pipe failure that can not be isolated.

ANSWER: The pool liner penetrations include: the inlet and outlet Primary Coolant pipinE, the pool cverflow weir drain, the pneumatic tube, and the beam tubes recirculation supply piping. A brcak in the Primary piping outside of the pool could occur by way of Primary Pump mechanical seal failure, Demineralizer Pump mechanical seal failure, Primary Heat Exchanger tube leak, or a breach of the Primary or Demineralizer System piping. Refer to SAR Figures 4-1A and 4-IF. Closing of the pool inlet and isolation valves at the pool boundary can be used to terminate the water loss from these specific avenues. The pool overflow weir and pneumatic tube enters the pool at or above the 20 feet level above the core and therefore are not considered a practical water loss path. The beam tube recirculation piping involves small bore pipes that draw pool water through the beam tube annulus and ultimately into the Primary Pump suction. A break in the recirculation piping beyond the biological shield could be isolated at the valve trench that surrounds the base of the biological shield. Refer to SAR Figures 4-1A and 4-1C for details of the beam tube recirculation piping. An

• 27'

NCSU PULSTAR December 15, 1989 Response to Request for Additional Information Revision 0 indirect method of losing pool water could occur with the failure of a beam tube that allowed pool water to enter the beam tube. In this case, pool water could migrate to the beam tube vent and reach the Bay floor at the water loop seal in the Beam Tube & Thermal Column Fan suction header. In this leak scenario,

-all the pertinent isolation valves on the affected beam tube would be closed to secure the leak path. Our response to questic6n # 61 deals with the i

consequences of a loss of pool water from a leak that cannot be isolated such that core uncovery occurs.

61. QUESTION: Section 13.2.1.4 - (a) Please provide more details about your use of the LPTR loss of coolant data in application to PULSTAR fuel. Are there more recent and more relevant data? (b) Please give more details about the I

direct and scattered radiation above the pool and near (or in) the control room.

Have you included back scattering from the roof? Where is the 250 mrem per L

hour location?

l L

ANSWER:

(a) In experiments at LPTR, water was removed from the system while the

' reactor was operating and thermocouples measured surface temperatures in the core following the loss of water. In all cases of tests, the maximum temperature was reached approximately 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after loss of water.

Specifically, during the 1 MW operation test, the driving AT between the maximum fuel surface temperature and the air was measured to be 398'F.

Hydraulic and heat transfer characteristics of the LPTR and the PUI. STAR are then compared to give an estimate of the maximum surface temperature in the PUI. STAR following the loss of water. The hydraulic diameter of the LPTR fuel was 0.0187 feet while the PUMTAR is 0.025 feet, and the heat transfer area of the IFfR core was 273 ft', while the i

PUUTAR core is 154 ft'. The following comparison is considere_d to be a good estimate of the maximum fuel pin surface temperature for the loss of pool water accident in the PULSTAR:

h = C, V"/D" and V = C, D" therefore, h = C, D" Where:

h = heat transfer coefficient V = velocity of coolant D = hydraulic diameter C, = constant

Then, h,w, = hm (Dm/D,oo)" = 1.06 hm AT,,, = ATm (h,,,/hm)(Am/A,

)

EL-

l

~ NCSU PULSTAR December 15,1989 Response to Request for Additional leformation Revisicn 0 AT,

= 398 (1.06)4 (273 /154)

AT,m, = 665'F It therefore indicates that the maximum fuel surface temperature rise should be on the order of 665'F, such that the absolute temperature should be around 765'F, far below the threshold for damage to the fuel or clad.

In regards to more recent analysis, a Master's Project by Mr. Jin-Wang Su was published in 1984 at North Carolina State University that dealt with the loss of coolant accident in a strict theoretical fashion. The results indicate that for a 48 hour5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> continuous run, the resulting maximum fuel average temperature would be on the order of.1040 F at the fuel mid position, with an associated air coolant temperature leaving the assembly of approximately 774'F, A-1.5 year continuous run would yield an approximate 1270*F maximum fuel average temperature at the fuel mid position and an air coolant temperature leaving the assembly of I

approximately 923'F. It should be noted that the NCSU PULSTAR has L

only operated for greater than 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> once in the facility's lifetime and occurred during a xenon equilibrium test (approximately 54 hours6.25e-4 days <br />0.015 hours <br />8.928571e-5 weeks <br />2.0547e-5 months <br />). The Master Project does not detail maximum fuel surface temperature that can I

be compared easily with the 765*F calculated from the LPTR test data, but rather the surface temperature lies somewhere between the fuel average temperature and the coolant temperature (both of which are a function of l

height) and is governed by the air boundary layer at the fuel surface. For the 1.5 year calculation, the maximum surface temperature lies between the mid position (where the fuel average temperature is approximately 1270*F and an air coolant temperature of approximately 800*F) and the top of the fuel assembly (where the average fuel temperature is approximately 1080 F and the air coolant temperature is approximately 962*F). As to why the theoretical calculation results in slightly higher value, this recent calculation takes no credit for any radial heat transfer between assembly boxes or to L

the external surfaces of the core that are surrounding by air. The L

calculation specifically assumes that all heat generated in the assembly is L

transferred only to the air inside that particular assembly. In addition it l

takes no credit for axial he 6 conduction transfer from the fuel assemblies H

to the grid plate or other metal structural support members. And finally, the calculation assumes an instantaneous loss of water, which in reality does take a finite time and will change the initial fuel temperatures, since decay heat is highest early in the transient. In any case, even this later more conservative calculation demonstrates that the fuel temperatures remain significantly below thresholds that would lead to either clad damage or fuel melting.

(b) The Monte Carlo analysis on the loss of pool water provided isodose lines based by simulated point detectors placed at several hundred locations in and around the reactor. The analysis allowed for both direct and scattered radiation from air and the surrounding structures (including walls, roof, floor, and biological shield). In addition, the energy loss of gammas penetrating concrete surfaces and then reflecting back out at lower energies was modeled. As indicated in the SAR section 13.2.1.4, it is estimated that the dose rate in the Control Room (i.e. standing at the window and l

' 29

. NCSU PUISTAR December 15,1989 Response to Request for Additional leformation Revision 0 looking through to the reactor) would be 250 13 mrem /hr at 10 minutes after reactor shutdown from loss of water. This dose results from gamma streaming through the wood core Bay door and the large plate glass win: low in the Control Room. The Control Room dose (at 10 minutes) decreases as one moves east from the windows to values of approximately 150 mrem /hr midway on the console,.50 mrem /hr at the east end of the console, and 10 - 25 mrem /hr at the steel door entrance to the Control Room. The highest calculated dose of 230 10 Rem /hr was found 20 foot directly above the pool (in line of sight of the core).

62. QUESTION: Section 13.2.1.5 - What would be the results of your fuel pin failure analysis assuming the maximum fuel burnup that may occur during the relicensing period?

ANSWER: Table 131. details the annuli activity in the core with all isotopes except "Cs and "Sr in saturation (which for calculational purposes, assumes 180 days continuous operation at 1 MW). Therefore, operating at full power continuously for 6,320 hours0.0037 days <br />0.0889 hours <br />5.291005e-4 weeks <br />1.2176e-4 months <br /> (equivalent to 20,000 MWD /MTU), will not increase the off site dose associated with the noble gases and halogens isctopes.

For *Cs and "Sr, annuli activity will increase if one assumes in the most conservative case that the operation is continuous for 6,320 days, and resalts in a revised Table 131 (included in Amendment 9) for annull activity that is released to the coolant. - As stated in SAR section 13.2.1.5, it is assumed that the solids remain in solution in the pool and thus do not contribute to off site doses. The revised SAR Table 13-1 assumes 6,320 days (equivalent to an average burnup of 20,000 MWD /MTU) continuous operation and was calculated using methodology described in the AMF Advanced Pulse Reactor document.

63. QUESTION: Section 13.2.1.5 - Fuel Pin Clad Failure - (a) What is the source of the information in Table 13-1 and 13 2? Discuss and compare with TID 14844, for example. What percentages of various radiologically important radionuclides migrate to and accumulate in the pellet-clad gap? (b) Instead of MPC, please convert these values into the likely doses to the most exposed person in the unrestricted area for the event you have postulated. Also, discuss potential exposures to staff within the reactor room. State these doses for the entire event for both whole body and thyroid. Clearly state and justify any assumptions used. (c) Page 9 - Please justify the use of 97 percent retention.

(d) Page 13 - Please provide the details of your assertions, especially for the sentence beginning: "An eight hour average...." Clearly state and justify any assumptions used.

ANSWER:

(a) The source of information for Tables 13-1 and 13-2 is the " Advanced Pulse Reactor (APR)" document by AMF Atomics 1963 (SAR reference 13-5).

Tables 17,18,20 and 21 contained on pages 74, 78, 80 and 83, respectively, were the source of Tables 13-1 and 13 2. The calculation for Tables 17,18, etc., is contained in Appendix C of APR document. The revised SAR Table 13-1 (reflecting maximum burnup proposed for fuel as required by SAR question # 62) calculates annuli activity assuming 180 days continuous operation that would increase the noble gases and halogens to their saturation values. "Cs and "Sr were calculated assuming 6320 30

NCSU PUESTAR ~

December 15, 1989 Response to Request for Additional Information Revision 0 days continuous full power operation.

TID-14844 and SAR reference 13 5 use the same methodology for calculating the time-dependent and saturation activities of the radiosiotopes i

of interest in the UO, matrix. Specifically the following equations are common to TID-14844 and SAR reference 13 5:

q,

= 0.865 x 10'(P)(7,)(1 - e*)

where:

q,

= amount of isotope i contained in the reactor at shutdown (Ci)

.P

= rated power (MW) y,_

= fission yield of isotope _i (atoms, fission)

A

=. decay constant for isotope i (sec')

and the saturation activity becomes:

q,

= 0.865 x 10' (P)(7.) in curies SAR reference 13-5 then provides additional methodology for calculating fission product migration to the annuli based on fuel temperatures tvolcal to the PULSTAR. The fraction of fission product activity that migrate to the gap are as follows (note that all values represent core totals):

Isotone Curies in Fuel Curies in Annu11 Fraction y

Kr-88 3.08 x 10' 0.787 2.55 x 10*

Xe-133 5.73 x 10' 11.3 1.97 x 10' Xe-138 4.96 x 10' O.404 8.15 x 104 I-131 2.68 x 10' 5.65 2.11 x 10" I-133 5.62 x 10' 6.81 1.21 x 10' Br-83 4.41 x 10*

11.9 2.70 x 104 Br-84 7.78 x 10*

0.10 1.28 x 10*

Cs 137 1.80 x 10' 53.0 2.94 x 104 l

Cs-138 4.97 x 10' O.55 1.11 x 10*

Cs-139 4.76 x 10' 23.2 4.87 x 10" l

Sr-89 4.14 x 10' 20.0 4.83 x 10" p

Sr-90

- 1.70 x 10' 44.4 2.61 x 104 l

Sr-91 5.02 x 10' O.75 1.49 x 10*

l There is no postulated scenario that leads to melting of NCSU PULSTAR l

fuel and therefore only annull activity is considered in the fission product release analysis. TID-14844 describes a typical PWR for which the maximum credible accident releases 100% noble gases,50% of the halogens, and 1% of the solids of the fission product inventory produced in 1-the UO, pellets. While this approach is appropriate to power reactors, where the need to analyze significant core melts exists (such as occurred at Three Mile Island), it is not, however, applicable to the NCSU PULSTAR.

Only annuli activity is considered in the failure of fuel pins at the PUISTAR.

(b) In order to fully respond to the whole body dose and thyroid dose at the 1

l 31

j 1

NCSU PULSTAR Decca ber 15,1989 Response to Request for Additional Information Revision 0 various locations requested, SAR Section 13.2.1.5 was significantly revised by Amendment 9.

Refer to this revised section for assumptions taken and the results concerning offsite exposure to a three fuel pin failure scenario.

Calculated potential exposures to personnel within the Reactor Bay from inhalation of the fission product gases, assuming 2.4 x 10' m' volume, no allowance for radioactive decay and a 2.35 hour4.050926e-4 days <br />0.00972 hours <br />5.787037e-5 weeks <br />1.33175e-5 months <br /> removal time, with continuous occupancy in the Bay, are as follows:

Quantity Activity Total Body Thyroid Isotope (Curies)

(uCi/mn Dose (mrem)

Dose (mrem)

Kr-88 3.78 x 10*

1.57 x 104 7.09 Xe-133 5.42 x 10*

2.26 x 10

7.91 Xe-138 1.94 x 10*

8.09 x 10

6.40 l

I131 8.14 x 10*

3.39 x 10

O.41 133 x 10

I-133

~ 9.80 x 10*

4.08 x 10

O.76 3.49 x 10' Br 83 1.72 x 10*

7.16 x 10

O.44 BR-84 1,43 x 10' 5.96 x 10*

0.05 (c)

The source of information is Table 4.4 in the " Fuel Defect Test - Borax IV" by R.F.S. Robertson et. al (ANI. 5862,1959). Table 4.4 shows only 3% of the radioiodines and bromines escape from the water, hence there is 97% retention in water (pages 14 and 22).

(d) The revised SAR Section 13.2.1.5 provides detailed assumptions on the calculation of dispersion factors and doses to offsite locations. Therefore, p

this specific question is no longer applicable.

64.

QUESTION: Section 13.2.2.1 - How is accidental removal or ejection of a control rod precluded?

ANSWER: Control rods cannot be removed without first removing adjacent fuel as described in the response to SAR question # 24. Since the standard control rods are on a mechanical lead screw with a drive motor and transmitter mounted above the lead screw, physical ejection of the standard control rod is not possible. The existing Pulse Rod air supply lines are to be removed and l

permanently capped to preclude accidental ejection of the Pulse Rod.

65. QUESTION: Appendix 3c, pages 9 Please provide a similar analysis with the following initial conditions: (a) Fuel at the normal temperatures for 1 hnV steady state operation, and 400 gpm coolant flow, and (b) Fuel at the normal temperatures for either 250 kW, or at ambient water temperature, but no coolant ' flow.

To answer the questio' a more detailed fuel calculation shall be ANSWER:

n derived to provide the initial conditions required in (a) and (b). Note that SAR Appendix 3A was revised by Amendment 9 to include this new analysis.

First, the following variables apply:

w

]

NCSU PUESTAR

~ December 15,1989

- Response is Request for Additional liforni: tion Revision 0 O',,, = average heat generated per ft of fuel rod = 2730.4 Btu /hr ft*F O',,, = maximum heat generated per ft of fuel rod = 2.92 O', = 7973 Btu /hr ft'F

r..= outer radius of fuel pellet = 0.2115 inches '= 0.017625 ft ri = inner radius of clad = 0.2157 inches = 0.01798 ft r, = outer radius of clad = 0'.2362 inches = 0.01969 ft k,,,,i = thermal conductivity of UO, fuel = 2 Btu /hr ft'F k,,, '= thermal conductMty of He fill gas = 0.1 Btu /hr ft*F ku = thermal conductivity of zircaloy clad = 8.2 Btu /hr-ft*F h, = heat transfer coefficient of coolant (values follow in analysis)

The value of UO, is a strong function of fuel temperature and ranges from 2 Btu /hr-ft*F at temperatures expected during a pulse to values over 3 Btu /hr-ft*F at the typical. steady state operating temperatures. To provide a conservative result, the lower value shall be assumed. The O', value would appear in the hot channel, where the total peak to average of 2.92 is expected to occur.

The centerline fuel temperature is given by the following terms:

T.,,,,,, = T

+ AT. + AT. + AT, + ATw i

l T,,,,,,,,, = T

+ [O'/n]{[1/2r,h,] + [(1/2ku)(In(r,/ri))] + [(1/2km)ln(ri/r.)] +

[1/4kw]}

. T,,,,,,,,, = T

, + [O'/2n]{[r,h,] +- [(ku)(In(r2/ri))] + [(km)In(ri/r.)] +

l

[2k,,,i]}

l L

For example, at 1 MW with a nominal flow of 500 gpm (yielding a h,=700 Btu /hr-ft' *F as detailed in Appendix 3B, page II V), the average fuel centerline temperature is calculated to be:

T,,,,,,,,, = 105'F + [(2730.4)/(6.283)){[(0.01969)(700)] + [(8.2)'in(0.2362/0.2157)]

+ [(0.1)in(0.2157/0.2115)] + [2(2)]}

T.,,,,,,,, = 105'F + 434.5{[0.07255] + [0.01108] + [0.1966] + [0.25]}

T,,,,,,,,, = 105'F + 31.5'F + 4.8'F + 85.4*F + 108.6"F T,,,,,,,,, = 335*F The average temperature across the fuel pellet is assumed to be 1/2 of the temperature rise across the fuel pellet since the temperature distribution is parabolic. Therefore one can readily calculate this value for the prescribed conditions:

T,,,i, = 105'F + 31.5'F + 4.8'F + 85.4*F + 0.5(108.6)'F T, = 281*F 33

W NCSU PUISTAR December 15, 1989

Response to Request for Additional Infor nation Revision 0 If the flow were to reduce to 80% (or 400 gpm), the heat transfer coefficient h, would be reduced by a factor of (0.8)", since h, is proportional to (velocity)".

This would yield a h, = 585 Btu /hr-ft' T. The revised average centerline temperature would be calculated as:

T

, = 1057 + [(2730.4)/(6.233)){[(0.01969)(585)] + [(8.2)In(0.2362/0.2157)]

l I

+ [(0.1)In(0.2157/0.2115)) + [2(2)]}

l T

, = 1057 + 37.7T + 4.87 + 85.47 + 108.67 1

T,,, = 3427 and calculating for the average temperature of the fuel pellet, l

Tw,,, = 287P_

For the maximum centerline fuel temperature at a nominal 1 MW and 100%

l flow, the value of O'w is used:

T,,

= 105T + [(7973)/(6.283)]{[(0.01969)(700)) + [(8.2)'In(0.2362/0.2157)]

+ [(0.1))in(0.2157/0.2115)) + [2(2))}

T,,,,,,, =.105T + 92.17 + 14.07 + 249.57 + 317.27 l

. T.,, = 7787 and calculating for' the cverage temperature of the fuel pellet, Tw, = 6197 Now in response to the specific questions regarding pulse performance, the initial temperatures are first calculated. Then, the adiabatic temperature rise calculated for a 58 MWsee pulse of 27007 (reference - Appendix 3C, page 11) is added to determine the maximum centerline fuel temperature during the pulse. This 27007 is the maximum allowed temperature rise at the hot spot and is derived from the hot spot energy density limit of 470 wattsec/gm assuming a 2.92 total peaking factor.

(a) For a nominal 1 MW and 80% flow condition the hot spot maximum fuel centerline temperature is calculated as:

T.,,,,,, = 1057 + [(7973)/(6.283)]([(0.01969)(585)] + [(8.2)'In(0.2362/0.2157))

+ [(0.1)]In(0.2157/0.2115)] + [2(2)]}

T

, = 1057 + 110.2T + 14.07 + 249.5T + 317.27 T,,,,,w = 7967 and calculating for the average temperature of the fuel pellet, Tw,,, = 637T If a 58 MWsee pulse occurred (with a total peaking factor of 2.92), the 34

NCSU PUIETAR December 15,1989 Response to Request for Additional Information Revision 0 new maximum local centerline temperature would become 796 + 2700 =

3496*F, which is well below the melting point of UO,.

It should be noted that while this calculation can easily be made, it does not represent a credible scenario. That is, the credible scenarios that could lead to an inadvertent transient (that could yield a 58 MWsec pulse) would occur from zero power and/or shutdown. For instance, fuel loading would certainly not be performed at 1 MW (our limiting reactivity accident) nor.would loading of high reactivity samples. Both of these exercises would be performed with the reactor shutdown. ' Furthermore, the analysis of the rod withdrawal accident has less severe consequences when initiated at high powers, and would result in a smaller energy release for inadvertent pulse (if not likely terminated before prompt criticality was even achieved).

(b) For a nominal 250 kw (25% power) and natural circulation, the AT. is presented in Table 1 of Appendix 3C. Using a AT. = 36'F at 250 kw, the initial fuel temperature is calculated as:

T

= 105'F + 36*F + [(1993)/(6.283)]([{8.2)'In(0.2362/0.2157)]

+ [(0.1)]'in(0.2157/0.2115)) + [2(2))}

T_ = 105*F + 36*F + 3.5'F + 62.4 F + 79.3*F T. = 286'F and calculating for the average temperature of the fuel pellet, T,,,,,

= 247*F i

If a 58 MWsec pulse occurred (with a total peaking factor of 2.92), the new maximum local centerline temperature would become 286 + 2700 =

2986'F, which is also well below the melting point of UO,.

1 l

1 l

35

i NCSU PULSTAR December 15, 1989 Response to Request f9r Additional Information Revision 0 j

NCSU ENVIRONMENTAL REPORT

\\

66. QUESTION: Page 2, Section 2, Paragraph 2 Give a quantitative comparison between the two cooling towers: (a) physical dimensions (b) heat dissipation i

capacity (c) use of chemicals to control corrosion and residue accumulation.

ANSWER:

(a) The PULSTAR tower is a Baltimore Aircoil, model VLT 200, with approximate dimensions of 128" high,145" wide and 57" deep. The BEL HVAC tower is a Baltimore Aircoil, model VLT 325, with approximate dimensions of 128" high,145" wide and 109" deep.

(b) The PULSTAR tower has a nominal rating of 200 tons, while the BEL HVAC tower has a nominal rating of 325 tons.

(c) The PULSTAR to'wer has historically used NALCO Chemical Company y

type 7374 Corrosion Inhibitor (which is chromate based). Approximately 5 gallons per year were used in the Secondary System. We are presently switching to NALCO 2820 (non-chrome /zine based). Sludge buildup in the PULSTAR tower is prevented by a typical blowdown scheme of ' fill and bleed" at approximately 5 gallons per minute. The BEL HVAC tower uses NALCO 2536 " Scale Remover and Inhibitor" in its fluid system. Note that the HVAC tower is under the complete control of the university Physical Plant (PP) organization, with. maintenance and repair provide by the PP.

67. QUESTION: Page 3, Section 2, Paragraph 2 - If not all of the radioactive gaseous effluent is Ar-41, what else is there? Give quantitative answers and L

I discuss.

1 ANSWER: In addition to the normal Ar release produced by reactor operation, the exhaust ventilation contains naturally occurring radon daughter products associated with the construction materials (i.e. granite, etc.) of the Reactor Bay. Their level, however, is quite low with respect to 'Ar production.

d

68. -QUESTION: Page 4, Section 4, Paragraph 2 - Give quantitative discussion of the changes in radioactive effluents when changing from air to nitrogen purging.

ANSWER: With the PULSTAR Reactor operating at 1 MW with the pneumatic system blower secured (i.e., air movement in the pneumatic tube is caused only by heating from the surrounding p'ool water), the Ar release at the stack top is measured to be 3.23 x 10* ci/cm. When the PN system is purged with nitrogen and other conditions remaining status quo, the ' Ar released at the stack decreases to a value of 7.2 x 10' pei/cm' (a reduction by a factor of 4.48).

69.

QUESTION: Page 5, Paragraph 1 - It is stated that federal regulations are met.

Please provide the concentrations of radioactivity released, and discuss your ALARA considerations. Quantities may be expressible as MPC-hrs / year.

ANSWER: The average concentration of liquid releases from the PULSTAR waste tanks (1975 to present) is 1.48 x 10* Ci/ml. This data is based on gross beta / gamma analysis, including tritium. The PULSTAR waste tanks system is 36

NCSU PULSTAR December 15, 1989 Response to Request for Additional Information Revision 0 comprised of three 904 gallon fiberglass tanks enclosed in a concrete vault approximately 6 feet below ground. This design provides isolation and shielding for personnel safety and AI. ARA considerations. Dose rates directly above the tanks have been measured with the tanks both full and empty, and typically are 15 7 20 pR/hr (the same as background for the surrounding area). Only two of the three waste tanks are kept on line for daily operation. The third tank is held for use in the event a volume of highly concentrated radioactive liquid requires longer storage to achieve optimal decay. This extra tank also provides a space for liquids during an emergency. The cycling of tanks and releases are strictly controlled by the Health Physics staff. All tanks and control valves are locked and require keys to gain access. Maximum dilution may also be achieved by releasing a volume of liquid over a long period of time. In a worst case scenario, the contents of a waste tank could conceivably be pumped into a transport vehicle for ultimate disposal.

Assuming the average Eampus sanitary sewer discharge of 100,000 gallons of water per day and the entire tank discharge concentration is tritium (as a limiting case):

1 1 05 gallons x 365 days = 3.65 x 105 gallons per year = 1.38 x 10" ml per year Applying the annual dilution to the average concentratiom 1.48 x 10' /1.38 x 10" = 1.07 x 10" Ci/ml Where the tritium MPC = 10' pCi/ml l

Assuming the reference man drinks 2.2 liters per 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, his daily consumption of this water would be 2.35 x 1048 Ci (or 9.80 x 10'"

Ci/hr), or in other terms, he would have to drink this water for 1.02 x 10* hours to r:ceive 1 MPC of tritium.

70.

QUESTION: Page 6, Section 6, Discuss storage and disposition of spent fuel.

ANSWER: Up to this point the PULSTAR Facility has not generated any spent fuel. We do however, have 8 unirradiated fuel assemblies in the pool fuel storage pits. At the time that spent fuel is removed from the core, they will be stored in the fuel storage pits to allow for a suitable decay. The storage pits together can hold 26 fuel assemblies (whereas our core arrangement holds 25 fuel assemblies). The fresh assemblies mentioned earlier, would then occupy positions on the core grid plate. After the decay, disposition of the fuel would take place by bringing in a shipping cask, lowering it into the pool, and then transferring the fuel into the cask, maintaining adequate shielding at all times.

The Reactor Bay crane, Bay mezzanine loading dock, and outside loading dock were all designed to accommodate a tractor-trailer shipment of spent fuel.

71.

QUESTION: You take credit for both 10,000 cfm and 12,500 cfm airflows.

Please discuss what provisions assure that both are operating when required.

ANSWER: The status of the Reactor Bay exhaust fan (10,050 cfm) is displayed on the Radiation Alarm Panel and the start-stop switch panel on the console, 37

+

f i

NCSU PULSTAR

- Daember 15,1989.

Resposw to Request for Additional Information Revision 0 l

and thus is easily discernible. The second flow rate of 12,500 cfm comes from f

the old wing of Burlington building ventilation system. This later system has redundant fans with an auto-start feature of the second fan if the first falls.

'One fan' operates at all times. We are however investigating a pressure switch or starter contact display signal that would be placed in the PULSTAR Control i

Room for further status indication of these fans.=

i i

t

+

n h

e 1

4 38

e 4

4 M NM December 15,1989 Response to Regnest for Additional Information Revision 0 TECHNICAL SPECIFICATIONS

72. QUESTION: Please submit revised technical specifications based upon the discussions that occurred during our site visit.

' ANSWER: Amendment 11 of the PULSTAR Technical Specifications is l'

enclosed, 4

4 4

l.

l.

l l.

i I

l' l.

l l

4 o

39

I NCSU PUIETAR December 15,1989

~ Response to Request for Additional Information Revision 0 REQUALIFICATION PROGRAM 73.- QUESTION: Section 8 The requirements of 10 CFR 55.59(c) (3) state that.

the requalification program must include on the job training addressing control manipulations. Please clarify which of the control manipulations addressed in 10 CFR 55.59(c)(3)(i) apply to your facility and include this list in Section 8.

ANSWER: Section 8 of the PULSTAR Requalification Program will be resised to reflect that our On the Job (OJT) Training includes the following items detailed in 10 CFR 55.59 (c)(3)(i): (A), (B), (E), (G)(2), (G)(3), (H), (I), (J),

t (M), (Q), (R), (S), (U), (W), (Y), and (AA).

74 QUESTION: Section 9 - The requirements of 10 CFR 55.59(a)(2)(ii) state that operators and senior operators must be examined such that they demonstrate an understanding of the ability to perform the actions necessary to accomplish a comprehensive sample of items specified in 10 CFR 55.45(a)(2) through (13) inclusive to the extent applicable to the facility. Please clarify which of these items are applicable to your facility and include them in the list of items to be included in the annual operating examination.

ANSWER: Section 9 of the PULSTAR Requalification Program will be resised to reflect that the annual operating examination shall include all items listed under 10 CFR 55.45(a) (2) through (13), since all are applicable to our facility.

l L

l l

l 40

4" g

s i TABLE 7-1 4

IWlONITOR DETECTOR ALERT ALARM DETECTOR TYPE COMMENTS LOCATION Control Room Above Control Room 2.5 mR/hr 25 mR/hr Dual Coaxial lon ALERT causes Mediation Alert" annunciation on north window Chamber Control Console. ALARM causes Evacuation.

Pool Reactor Bridge to mR/hr 100 mR/hr -

Dual Coaxial lon ALERT causes " Radiation Alert" annunciation on Chamber Control Console. ALARM causes Evacuatiort

'I West Wall West Reector Bay 10 mR/hr 100 mR/hr Dual Coaxial lon ALERT causes Padiation Alert" annunciation on.

Wall Chamber Control Console. ALARM causes Evacuation.

Primary Demin MER Mounted at 100 mR/hr 125 mR/hr Dual Coaxial lon Nert causes Tadiation Alert

• annunciation. Mso Demineralizer Chamber has local readout and alarm.

VAMP Reactor Bridge (N/A) 2.5 mR/hr GM Tube Local Auditde and Visual Aiamt Stack Gas Stack Sampling Unit 500 cpm 3 x 10' cpm GM Tube Algt cause Padiation Alert

• annunciation on Control Console. M causes Evacuation..

Particu! ate Stack Sampling Unit 10' cpm 10' cpm Beta Scintillator Alert cause Padiation Alert" annunciation on Control Console. Nqwm causes Evacuattort Auxiliary GM Exhaust Duct 2 x 10' cpm 8 x 10' cpm

. GM Tube Alert cause Padiation Alert" annunciation on -

Control Console. Alarm causes Evacuatiort FilterGM Exhaust Filter Bank 6 x 10' cpm 4 x 10* cpm GM Tube Alert causes Tadiation Alert" annunciation on Control Console.

4 CAM Reactor Bridge (N/A) 4.5 x 10' cpm Beta ScintMator Local Audible and Visual Alemt t

.C Waste Tanks #1 Dry Well!n Each Tank 8 x 10' cpm 8 x 10' cpm Gamma Scintillator.

No console annunciatiort

1. 2, & #3 j

I e

i I

.. ~..,

s.-...

4 Revision 2 March 1,1989 HP Procedure 20-2 j

Release of Radioactive Wastes to the Sanitary Sewer System

1.0 OBJECTIVE

To provide guidance for analysis and dilution of low level radioactive liquid wastes in order to comply with 10 CFR 20 regarding such release to the i

sanitary sewer system, j

2 c0 RE1ATED PROCEDURES:

HP 20 3 HP-20 4 HP 20 5 HP 20-6 HP 20 7

3.0 REFERENCES

i 10 CFR 20. Title 10, Code of Federal Regulations, Part 20.303 Final Safety Analysis Report, Section 10.

i i

4.0 DISCUSSION

l' The Assistant Director for Operating Reactors, NRC, authorized NCSU, by letter dated 17 May 1973 to the Radiological Protection Officer, the use of "the minimum average daily water input to the campus...as that quantity of dilutant for determining that the average concentration is equal to the limits specified in Appendix B, Table 1, Column 2 of 10 CFR 20..." The specific amount of diluent that can be used is not more than 671,000 gallons of water per day.

The worst situation would be the three PULSTAR holding tanks completely filled with radioactive vastes.

In this situation, the maximum specific l

t l

1

HP 20 2 Revision 2 Release of Radioactive Water to the Sanitary Sewer Syttem March 1,1989 activity the above amount of diluent would accommodate, assuming all wastes d

were to be released within the same 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period, is 1.0 x 10 pCi/ml.

-5.0 PROCEDURES:

The following guidelines apply to the release of low level radioactive liquid wastes from the Department of Nuclear Engineering holding tanks (three PULSTAR tanks):

Specific Activity Action 4 x 10'? pCi/ml Gross $ + y count, tritium and Samma analysis.

Release and record.

4 4 x 10 pCi/ml Gross p + y count, tritium and gamma analysis.

Release and record.

4 x 10-5 pCi/ml Gross $ + y count, tritium and gamma analysis, consults with RSO prior to relcase.

The above guidelines assume there is no other germane information that would prevent the release of these vastes.

PULSTAR Vaste Holding Tanks 3 Tanks, each 904 gallons - 2,712 gallons Campus Dilution - 671,000 gallons 2

p-HP 20-2 Revision 2 Release of Radioactive Water to the Sanitary Sewer System March 1,1989

~ Calculating for maximum tank specific activity (pC1/ml) that could be released without exceeding 4 x 10'7pCi/ml (10 CTR 20):

pCiman (4 x 107 pC1/ml) (671,000 + 2,712 gal.)

~

al (2,712 gal.)

pCiman 9.92 x 10'8 pCi/ml

,1

6.0 APPROVALS

Associate Director:

• NT

/

s M }[

/fE)

Date:

Mod (QLLECh Ghn Department Head:

j Date:

M O9 N i

Radiation Protection Council:

1. 65% N'/ b nI-7[A0/lf Date:

e 3

Revision 2 March 1,1989 HP Procedure 20 3 Samolina Waste Water Tanks i

1.0 OBJEcTIVEr To provide guidance for vaste water tanks sampling in cortpliance with 10 CTR 20 reletive to release of lov level radioactive vaste to the sanitary sever system.

2.0 RELA *1ED PROCEDURES:

l

}{P 20 2 l

lip.20 4 MP 20 5 AP.20 6 ilP 20 7

3.0 RETERENCES

l 10 CFR 20. Title 10. Code of Federal Regulations, Part 20 FSAR, Final Safety Analysis Rrpcrt, Section 10.

4.0 DISCUSSION

It'e PULSTAR racility has three 904 gallon tanks into which the low i

level liquid vaste is collected. These tanks are monitored and analyzed for radioactivity prior to release into the sanitary seser system to be certain of compliance with 10 CTR 20 at.d other regulations and requirements.

Liquid volumes are required to be turned over twice to insure proper mixing before sampling.

1 1

j

[

HP 20-3 Revision 2 Sampling Weste Water Tanks March 1,1989

)

5.0 PROCEDURES

(1)

Equipment:

Electric pump with hoses and tap for bottle fill Liter bottles marked for identification of tank number Extension cord Plastic bags Plastic gloves Paper towels Absorbent paper Cart to move equipment Vent cap keys (2)

Prepare bottles in accordance with standard laboratory practices.

(3)

Check sump tank in TULSTAR Mechanical Equipment Room.

If this te.nk is at least 2/3 filled, contact Operations and pump out.

(4)

Close inlet valves (located in B103) on tanks to be sanpled.

(5)

Open inlet valve on empty tank and check that outlet valve is closed.

(6)

Ensure that switch on pump cart is OTT (handle down) then connect receptacle end of extension cord into switch box plug on cart.

(7)

Connect extension cord (plug end) into receptacle located on the outside of the southwest corner of Burlington Laboratories.

(8)

Connect pump intake hose to extension pipe extending from tank bottom to vent cap.

This extension pipe will be located in vent pipe of tank from last sampling (check discharge record for the tank last sampled.)

2

r HP 243 Redslon 2 Sampling Weste Water Tanks Match 1,1989 (9)

Place the pump discharge hose into vent pipe of empty tank (at least 2 feet of hose should be down in pipe.)

(10)

Turn pump switch on QH and allow =1 minute for pump to pick up a full head of water. The height of water can be monitored through the clear intake hose.

NOTE:

Pump is capable of pumping 80 gallons a minute, usually 9 to 13 minutes to pump tank depending upon how full the tank is.

(11) When the tank is pumped empty, turn switch DIE and reverse intake and discharge hoses, pump volume back into the tank previously pumped from.

(12)

Based upon time required to pump volume from original tank (first pump), take sample at end of second pumping.

(13)

Till sample bottle from pump tap over absorbent paper.

(14)

Sample all subsequent tanks using same procedure (turn tank l

l volume over at least twice.)

(15) At completion of tank sampling, disconnect extension cord from receptacle first, then coil on top of care for storing.

(16)

Coil pump hoses on each side of pump between upper and lower cart shelves.

(17)

Ensure vent caps are locked and keys are returned to lab.

J (18) Measure the depth of water in the tank with the dipstick (2 meter stick) located under EPA lab window. Take reading in inches for FULSTAR tanks.

]

l 3

HP 20-3 Revision 2 Sampling Weste Water Tanks March 1,1989 (19) Clean up area, take in solid waste. DO NOT forget to replace neter stick under EPA vindow.

CAUTION: If the sanple is for NAA, acidify upon returning to the laboratory, i

6.0 APPROVA13t Associate l Director:

- j d Y-6

/

Date:

Es. ' S}' /?d9 Departn,ent Head:

We

/

oh Date:

N Q9 N Radiation Protection Council:

44e M /IM2s

/

,7 M / f f Date:

e 4

i i

nge-oooooooooooooo S

ngoooooooooooooooo 4

t

-Roooooooooooooooo S

ERaooooooooooooooo e

g gg-*co-ooooooooooo, 8

3 genooooooooooooooo 8

i I

N p3wooooooooooooooo gg-ooooooooooooooo a

bg-ooooooooooooooo h

s S

w3gwoooooooooooooo g

ng:-oooooooooooooo k

I ntyn-ooooooooooooo Y

7 l

s g

eggoo-oooooooooooo 8

t I

E wp*ooooooooooooooo l

9 e.n~oooooooooooooo e

oggNeooooooooooooo n

@ nueooo.cocooooooo.

1 t

9 1

li i

Innes i

Iy eod on.oosome*.

SSSS

$G999922993S9*

unemesoot:0

1 North Carolina State University Raleigh, North Carolina November 1,1989 P

ALARA Philosophy Statement North Carolina State University is committed to a policy of making every effort to keep radiation exposures as far below the regulatory limits as reasonably achievable, nus, the underlying philosophy of the Radiation Protection Office, Radiation Protection Council and particularly the PULSTAR Reactor Facility, will be to maintain radiation exposures "as low as reasonably achievable." His philosophy

~

referred to as ALARA is in keeping with the recommendations of the National l

Council on Radiation Protection and Measurements, the National Academy of Sciences National Research Council, and other independent scientific organizations.

The principle of AIARA is also codified as part of the Nuclear Regulatory Commission regulations In Section 20.1 (c) of Title 10, Part 20 Code of Federal Regulations, which states that licensees should "make every reasonable effort to maintain radiation exposures, and releases of radioactive materials in effluents to unrestricted areas, as low as is reasonably achievable."

r

~Associat(' Director of the Nuclear Reactor Program W !U Wk%W Radiation Protectiopfficer

?uAoud % $-

C rjnan of the Radiation Plotection' Council l

Ace Chancellor [

~

l l

3