Saturday, May 31, 2008

UHI and RELAP 2008

Here is an abstract of a Power Point presentation that I want to present at the 2008 RELAP International Users Seminar. The first item is an e-mail from Dr. Dean Dobranich of Sandia that has brief comments related to the abstract. The next item is the abstract. I have not heard back from INL regarding my request for space on the agenda.


RE: 20+ years ago
6/29/06 1:36:10 PM Mountain Daylight Time
Reply To:



Nice job. My only comment concerns the last sentence. I suggest putting the emphasis on the fact that analysis is warranted to investigate the scenario you suggest. The way I read the sentence now, it seems like the emphasis is on using SCDAP/RELAP5-3D as if you are trying to drum up business for that code package. A possible alternative: "Additional UHI scenarios of concern should be investigated, perhaps using SCDAP/RELAP5-3D."

Dean Dobranich, PhD, DMTS Thermal and Reactive Processes, 1516 SNL MS 0836, Voice: (505) 845-0446, Fax: 844-8251
From: [] Sent: Thursday, June 29, 2006 1:12 PMTo: Dobranich, DeanSubject: 20+ years ago

Thank you for your searches, etc. The attached is the abstract of a PowerPoint presentation that I am preparing for an August RELAP meeting at West Yellowstone. There is maybe a 30% chance that this will be allowed.



Clarifying History via an Application of SCDAP/RELAP5-3D: Analysis of Severe Accidents for Pressurized Water Reactors with Degraded Upper Head Injection

Robert H. Leyse, CEO
Inz, Inc.

During the early 1980s the Upper Head Injection System (UHI) was added to the Emergency Core Cooling Systems (ECCS) of several pressurized water reactors having Ice Condenser Containments. UHI was added to regain operating flexibility that was lost due to the impact of the Ice Condenser design upon the 10CFR50.46 Appendix K ECCS Evaluation Model Analyses. Ice condenser design resulted in a reduced containment back pressure during a design basis loss of coolant accident decreasing the amount of steam that could be vented to the containment. The ECCS analysis results therefore required more limiting restrictions on normal plant operations in order to satisfy Appendix K requirements than were required for dry containments.

The UHI system was a high-pressure emergency injection system which supplemented the conventional ECCS. It consisted of two 1800 cubic foot accumulator tanks, one filled with subcooled borated water and the other with compressed nitrogen at 1250 psia. They were separated by a membrane that prevented pressurized nitrogen from dissolving in the borated water. This membrane ruptures once the water-filled accumulator begins delivery. As the reactor system pressure falls below 1250 psia during a loss of coolant accident (LOCA), passive check valves open and UHI borated water enters the upper head of the reactor vessel. After 1000 cubic feet of borated water has been injected, isolation valves close on a low level trip and injection ceases.

UHI systems were placed in operation during the early 1980s. The systems proved troublesome and the NRC evaluated the operating experience (reference 1), however, the NRC analysts did not identify any safety concerns. Then, on October 31, 1984, Duke Power Company shut down its McGuire Unit 1 when it determined (reference 2) that the UHI level transmitters were improperly installed and the UHI valves would have failed to isolate in the event of a LOCA.

Duke Power Company elected to remove UHI from its ice condenser plants and submitted its proposed Technical Specification Revision (reference 3) to the NRC. Also, the NRC commissioned Sandia National Laboratories to calculate the impact of failure to isolate the UHI on core temperatures during a LOCA.

On January 29, 1986, Dr. Dean Dobranich of Sandia commented in its letter report (reference 4), “These calculations suggest that failure of the UHI shutoff valves to close during a large-break LOCA may not be detrimental to the cooling of the core. The extra water injected by the upper head accumulator provides additional core cooling and the injected nitrogen flows out of the primary system without significantly interfering with refill and reflood.”

The NRC negligently (criminally) confined the Sandia calculations to the stuck-valve condition. The NRC did not include the case of a long term ruptured diaphragm prior to the LOCA, in which case the injected water would carry a substantial amount of dissolved nitrogen. More seriously, the NRC did not consider the situation of improperly functioning level transmitters, in which case there was no assurance that the water-filled accumulator had 1800 cubic feet of water, or any water. Clearly, the setup for a fast moving severe accident of Chernobyl magnitude could have been in place at any of the six ice condenser plants with UHI. Additional UHI scenarios of concern should be investigated, perhaps using SCDAP/RELAP5-3D

1. NRC Report AEOD/E404, Failures in the Upper Head Injection System, February 28, 1984.
2. NRC Report of UHI Incident, Preliminary Report of Unusual Occurrence, PNO-II-84-81, November 2, 1984.
3. Letter from H. B. Tucker (Duke Power) to H. R. Denton (USNRC), Proposed Technical Specification Revision: Deletion of Upper Head Injection System, May 9, 1985.
4. Letter Report, Dobranich (Sandia) to Watt (USNRC), LOCA Calculations for UHI Plants, January 29, 1986.

Thursday, May 22, 2008

Closed Meeting at ACRS: Future Plant Designs

Following is an e-mail from NRC to Leyse

Closed ACRS Subcommittee Meeting - February 6, 2008
Date: 5/20/2008 8:45:32 A.M. Mountain Daylight Time




Mr. Leyse,

This is a response to your March 27, 2008 email regarding the closed ACRS Subcommittee meeting on February 6, 2008. This meeting of the Future Plant Designs Subcommittee was closed to prevent the premature disclosure of information which would be likely to significantly frustrate implementation of a proposed agency action per 5 U.S.C. 552b (c) (9) (B). The ACRS subcommittee heard from representatives of the Nuclear Regulatory Commission and Department of Energy regarding the proposed licensing strategy for the Next Generation Nuclear Plant (NGNP).

Cayetano Santos,
Reactor Safety Branch,

Tuesday, May 13, 2008

More on UHI (Upper Head Injection) Ultra High Risk

Of course, this is history. UHI has been out of the picture for some time.

You may look at my earlier entry of today. This is part of a very long package that I'll stitch together from this entry and much more to come. The first slide is a repeat from my earlier entry today. It is followed by slides that illustrate that years later, that reasoning is followed, namely that since UHI is not needed to handle a LOCA, the presence of UHI may be ignored in PRA's. In other words, you could store tons of explosives at a power reactor site and that would be OK because it is not needed to deal with a reactor accident.

I wrote my warning in October, 1994. The PRA for Sequoyah is dated January 1991. The last slide is copied from the Japan Atomic Industrial Forum (1990) and reveals plans to get rid of UHI at the two Ohi units. So, here are the slides. Click on the slide to enlarge and then click on the back arrow to get to the next slide.

The above EPRI memo of October 1994, implies that it is reasonable to ignore the presence of UHI in the McGuire PRA.

The above identifies the Sequoyah PRA dated January 1991, over five years later than the EPRI memo.

The above has a line that states UHI is not considered in this PRA. It thus copies the McGuire PRA.

The above has a paragraph that presents the weak argument for not including UHI in the PRA.

The above, November 1990, discusses the first steps to get rid of UHI at Ohi Units 1 and 2.

UHI (Upper Head Injection) Ultra High Risk and a Fantastic Coincidence

This is a brief logging of three documents. I've discussed UHI at least twice before. It will become a very lengthy package. For now here is a nutshell that documents a fantastic coincidence. I submitted my memo on October 3, 1984. In response, EPRI contacted Duke Power and Mr. Sugnet of EPRI briefly documented his discussions on October 25, 1984. Then it just so happened that on October 31, 1984, Duke found four out of four level detectors reverse connected at McGuire Unit 1. The NRC publicly reported this on November 2, 1984.

Indeed, this is a strange coincidence since the detectors had been reverse connected for a very long time. The NRC fined Duke $50,000. Duke argued that they should not be fined since they had discovered the fault independently. The NRC kept the fine at the outrageously low level of $50,000 because Duke had found and reported the deficiency without NRC involvement.

Here are the documents. Click on the document to enlarge and then click on your back arrow to get to the next document.

Friday, May 9, 2008

43 years ago Leyse developed sodium loops for GETR (Click on slides to enlarge)

I did great work in getting this going and I set up a fantastic 7-rod mockup with somewhat superb electrically heated fuel rod simulators . The racketeers got fed bucks to try to build more development rigs after I was fired. However, they had no luck in getting performance of the fuel rod simulators in sodium (high power heaters running at one kilowatt per inch at one quarter inch diameter).

I had written a report that described how I built and operated my mockup. I did well, but the rig needed further development. I had very few dollars and scrounged a lot. I did not even take time or money to procure the necessary swaging dies and used what the metallurgists had on the shelf. I recall using a 0.22 inch die for swaging the o.25 inch solid assembly, but I had luck and the high power heaters lasted long enough to collect a set of quality data at full power.

Well, when the chief GEnius took over, he had his GEnii follow my report to the letter. If they had looked into the fundamentals of swaging, they might have succeeded. Well, the clown has croaked, so I have no chance to needle him. Actually, I telephoned his place a long time ago, July 20, 1992, to really lay into him, but he had already permanently left the scene.

The last of the following six slides is likely of some relevance today because it has an impressive list of isotopes for sale. Again, click on the slides for enlargement, then click on your back arrow to get to the next slide. Or, skip it all and move on.

Following is one of the briefs that I wrote as I tried to get this going. And there will be more on this.

Of course, I'm working on this section, meaning that better stuff is on the way. This GE marketing brochure was on the street about 43 years ago. It is a good thing that no sodium loops were ever installed in the GETR.

Tuesday, May 6, 2008

Single plate burnout at GETR and Single Element Meltdown at WTR

GETR, TR-1, APED-4454 and WTR, TR-2, WTR-49

GETR means General Electric Test Reactor.
TR-1 means Test Reactor 1 license issued by Atomic Energy Commission (AEC) for GETR.
APED-4454 is a report, Atomic Power Equipment Department for the GETR single plate burnout.

WTR means Westinghouse Testing Reactor.
TR-2 means Test Reactor 2 license issued by AEC for WTR.
WTR-49 is a report issued by Westinghouse for WTR for single element meltdown.

I'm working on this. WTR had more engineering but GETR had more good luck! I'll discuss GETR first. GE burned out Element F-21 and did not even know it at the time, but I must disclose that Mr. Paul Zeanah, the GETR analyst during 1960, is the person who suggested that I inspect F-21 when I was troubleshooting the abrasion situation.

I selected the following set of 13 slides from the report of my work (APED 4454, Class III) and it has the handwritten suggestions by Dorothy Crosier, the GE editor. Please click on the following slides for enlargement. Then use your back arrow to move to the next slide.

I used the Bernath equation for calculating the burnout heat flux. GE had published a burnout correlation for this fuel when they were paid to design the Engineering Test Reactor (ETR). That GE correlation is useless (just plain wrong) and I also did not reference it when I proposed the power level increase for the GETR.
Next, let's look at the meltdown of one fuel element at the Westinghouse Testing Reactor (WTR). Following are five slides from report WTR-49. Les Kornblith, my boss at GE, gave me this report when he left GE to manage reactor inspectors at the AEC. (Les was a good man, hired into GE by Sam Untermyer, and Les did well with the feds.)

I'll add more slides from WTR-49 and further discussion at my leisure. For now, I'll say that GE had a far superior plant arrangement and they spent a lot less money. In Simpson's book, Nuclear Power from Underseas to Outer Space, he reports in 1995 that further cleanup from the meltdown of the single fuel element will be expensive, "...during the cleanup process following the accident thousands of gallons of water were used that became contaminated with radioactive water. Some of the water has contaminated the soil and groundwater. This cleanup must be completed before final decommissioning can be completed and will be quite costly."
Simpson also reported, "Unfortunately , shortly after it was completed, the government, despite having uged us to build the WTR, built two other testing reactors. The AEC built the Engineering Test Reactor (ETR) and NASA built one at Plum Brook in Ohio. This took away most of our customers causing the WTR to be uneconomic and it was retired in March 1962 ..."
In its analysis of the fuel element failure, the Westinghouse team did not use the Bernath correlation as I did for the GETR case. They used two equally applicable correlations: one from DuPont, the Mirshak equation, reported in DP-355, and another from Argonne, the Jens and Lottes correlation. Their burnout margins were greater than a factor of two for fuel as designed, as was the case for GETR.
There have been approaches to reactor plant licensing that have included the concept of the maximum credible accident. Back in the old days GE had an Organization and Policy Guide (OPG-24.1) that called for "...careful, comprehensive and imaginative ..." analyses of the mechanisms and consequences of reactor accidents. I'm not certain that I'll ever find my copy, but it is around. I am developing a new approach, called MAXIMUM POSSIBLE BLAST. I have never been an insider at WTR, but my feeling is that WTR very likely never even came close to having the potential to match GETR for the title of KING OF THE MAXIMUM POSSIBLE BLAST. Anyway, GETR was permanently shut down based on seismic stuff during 1977, over 15 years later than WTR.