Considerable advantages accrued to my joining this expertteam. Firstly I know the members personally and have discussedreactor safety frequently with them and in particular havediscussed safety of the reactors designed in the USSR. Thisenabled me to be aware of issues that otherwise I might havemissed. Secondly the Armenian government had ensured that therewere available to us all the most important people connected withthe plant recommissioning and operation, enabling me to makeenquiries and gain impressions which otherwise would not have beenpossible. I am grateful to the European Bank and to my colleaguesfor allowing me to accompany them on this visit. Although Ibelieve that much, if not all, that is said in this report willfind agreement among these experts, I have written this reportindependently and they are in no way responsible for the opinionspresented herein.
I am particularly grateful to the Deputy Prime Minister MrChichechian for welcoming me (and my wife who accompanied me) toArmenia as a part of the team.
Introductory History
According to the records of the International Atomic EnergyAgency (IAEA) the Medzamor nuclear power plants were begun in 1973and began commercial operation in 1979 and 1980 respectively. In1982 one unit was damaged in a fire, although no dangerousradioactivity escaped. After that fire, extensive modificationswere made to the plant to make it more secure against fire, and itwas reported by the IAEA that it was the best protected againstfire of all the VVER 440/230 reactors. In 1988, Armenia suffereda severe earthquake which caused extensive damage to nearby cities,although the nuclear power plant was undamaged. Nonetheless,according to reports at the time, the central committee of theCommunist party worried about the Medzamor plant immediately, asdid many others. Three ethnic Armenian engineers from BechtelCorporation of San Francisco, USA, visited the plant and maderecommendations to improve the earthquake resistance of auxiliaryequipment. After consideration of these and other matters, it wasdecided in Moscow, with the approval of Yerevan authorities, toshut the plant down. This decision seemed sensible to them at thetime, but geopolitical considerations make it seem less sensible inretrospect.
The energy problems of Armenia
Armenia has few indigenous fuel resources. In 1988 theelectrical generating capacity was about 3,500 MW(e) made up of 900Mw(e) nuclear, 1,750 MW(e) fossil fuel generators, and 900 MW(e)hydropower, mostly from Lake Sevan. In 1995, fuel shortages limitthe fossil fuel capacity to only 500 MW(e), and ecologicalconsiderations limit the hydropower from Lake Sevan to 400 MW(e)for a total of 900 MW(e). There are few other hydroelectricresources. Electricity is cut off to residential customers on arotating basis, and there is limited electricity even to run pumpsto supply water to the population.
In 1989, it was anticipated that Armenian energy policy couldbe part of a regional Caucasian energy policy. There is plentiful,moderately priced, natural gas in Turkmenistan, with an adequatepipe line through Azerbaijan. The electricity grid of the USSRseemed adequate and could be extended. In 1991 this changed. Thecivil war in Nogorny Karabagh led Azerbaijan and Turkey to lay aneconomic and fuel embargo on Armenia and there is no sign that itwill be lifted soon.
Fuel embargoes have been seen before in the world. Japanjoined Russia, Great Britain and France in the Great War of 1914 togain control of Manchurian coal which had been a German concession,and attacked Pearl Harbor in 1941 because of the oil embargo placedon Japan by Holland and the United States. History tells us thata country which has an inadequate fuel supply can be a desperateand dangerous country. Therefore it is important for peace in theCaucasus, and hence the world, that Armenia should have a degreeof short term energy independence even if the present embargo islifted. This leads to the conclusion that nuclear power forArmenia is essential both in the short term and in the long term. Armenia wishes not only to restart the Medzamor plant as soon aspossible, but to build new facilities to replace, and perhapsincrease, them. We were told orally that the Medzamor plant willoperate only seven more years, but this is clearly a matter ofdesire rather than of detailed planning. There is some plan for areplacement about 2005 but this will clearly need considerableinvestment. No sensible alternative to nuclear power, and to thetemporary operation of the Medzamor plant has been suggested byanyone. These two reasons, humanitarian aid to the Armenian peopleand help in securing the peace of the world were both of greatinterest to Andrei Dmitreyvich Sakharov during his last years, andare therefore both important to the Andrei Sakharov Foundation. However the power plants must be made as safe as reasonablypossible. This visit was to report on the efforts of the Armeniansto this end.
The Armenian Plan
Everyone agrees that the decisions about energy policy inArmenia and nuclear power in particular must now be made by theArmenians themselves. This includes scientists, engineers andpoliticians in Russia as well as western scientists. HoweverRussia and other countries can help. The government of Armeniadecided in 1993, with an official start in February 1994, torestart the Medzamor power plant. The electricity shortages mighthave suggested an immediate restart, but the decision was made tocheck every piece of equipment as if the plant was being startedfor the first time, and to add all the safety precautions that thebest engineers and scientists in the world could suggest. Specifically, it is to be as safe as any other power plant of thistype. This has entailed detailed work for the last 18 months, andthe results of much of this work are now visible. In this work thehelp of Russian engineers, and financial aid from the government ofthe Russian federation , has been essential and is much appreciatedby all Armenians. This work is expected to be finished by Octoberfor a possible restart on November 1st in time to help allay thefuel shortages in the forthcoming winter. However, the decisionhas been made to make safety an overriding priority so that if allthe additional safety features are not ready at this time restartwill be postponed.
Other Reactors of This Type
The Medzamor reactors are of the pressurized water type andare designated VVER 440/270. They are a slight improvement overthe VVER 440/230 (they have additional pumps for decay heat removalin the event of a severe earthquake) but the improvements are notas great as those the VVER 440/213. There are 10 reactors of thistype operating in NuovoVoronezh and Tula in Russia, Kozloduy inBulgaria, and Bohunice in Slovakia. Five plants in Greifswald,East Germany, were shut down soon after German reunification. Thesepressurized water reactors are very different from the RBMKreactors of which one exploded at Chernobyl and more similar towestern reactors. But they are still an old type which would notbe licensable in western Europe or the USA. Several reactors of aslightly improved type, VVER 440/213 are also operating in Russiaand Eastern Europe. As of 1989, all of the VVER 440/230s possessedthe following major deficiencies:
- Inadequate fire protectionInadequate earthquake protectionInadequate Emergency Core CoolingInsufficient redundancy for auxiliary equipmentNo containment vessel to contain the steam and fissionproducts in the event of a failure of decay heat removal.
It is not the purpose of this report to describe in detail thesafety analyses and deficiencies of this type. Nor to describe indetail the particular Armenian reactors. There are several reportsdescribing the Armenian from Bechtel corporation in USA, KurchatovInstitute and MINATOM in Russia, Framatome in France, the WorldAssociation of Nuclear Operators (WANO) and the InternationalAtomic Energy Agency (IAEA). However, a few general observationsmay be useful.
A pressurized water reactor needs a good reactor vessel inwhich the reactor and its coolant water are placed under pressure. This vessel will inevitably be irradiated with fast neutrons fromthe nuclear fission and this will make the vessel, and any welds,somewhat more brittle. The early Russian built reactor vesselswere particularly bad in this regard, but the Russians developedtechniques of measuring the brittleness of the welds and ofannealing them. Tests on the vessel on unit 2 suggest that it willbe safe for 10 to 20 more years.
The primary safety concern with pressurized water reactors isthat the heat be removed from the reactor core at all times. Evenafter shut down of the nuclear reaction there is decay heat thatmust be removed. There are several possible causes (initiators) oftrouble. If there is a fire or earthquake that shuts down thereactor, disconnects it from the electricity grid, and alsoprevents the start of emergency pumps, the reactor will heat up. With reactors in the USA fuel can start to melt (as it did at ThreeMile Island in 1989) within a couple of hours. However, in theVVER 440 series, the reactor can be isolated for at least 8 hoursbefore problems begin. This time can enable the staff to recoverdamaged equipment and regain control. This was proven in the 1982fire at Medzamor and at other VVER 440/230 reactors. This providesan inherent degree of safety that was not initially realized bymany scientists and engineers in the west. Nonetheless theauxiliary equipment must be fire and earthquake resistant.
The initial inadequate resistance of this and other VVER 440reactors to earthquake before 1990 can be illustrated by twodeficiencies. Relays and other control equipment are operated byemergency DC power provided by a set of batteries that arecontinually recharged. Before 1990 these were lead-acidaccumulators in glass jars, connected in series. If just one ofthese had shattered in an earthquake there would have been no DCpower. In 1995 these glass batteries have been replaced by a morerobust set of batteries bought in Germany.
If the steam turbines fail in an earthquake, and the coolingtowers fall down, they cannot remove heat from the reactor. In1988 there would be nothing to prevent ultimate melting of thereactor core. Now a cooling pond is being installed that willremain intact in a severe earthquake and which has adequatecapacity to cope with the shut down heat. The Medzamor reactorswill not restart until this cooling pond is finished.
The fire resistance has been much upgraded. 70 km of newcontrol cable has been installed and the routing of these cablesadjusted to assure independence of control in the event of fire. I saw many fire extinguishers which were absent in many Sovietlaboratories and these seemed to be in working order.
There is one small detail that gives me small pause. Thefloor of many of the passageways is covered with plastic. This isa feature common to many Soviet reactors and laboratories. I wasunable to discern whether the plastic is fire resistant or whetherit has been tested. I remember that in describing the Chernobylaccident the General in charge of the fire brigade described ingreat detail the presence of materials that were supposed to benon-inflammable but nonetheless burnt merrily.
Sabotage and Vulnerability
It has been suggested that the power plant is unusuallyattractive for a saboteur or terrorist from an unfriendly country,and that this possibility is, in itself, a reason for refrainingfrom operation. In western plants the containment vessel protectsmany parts of the plant from attack by a mortar shell. The VVER440/230 reactors have no containment and are therefore a littlemore vulnerable. However, an attack is more likely to come from aneighbor than from a distant country, and although the neighborwould be far less affected by any radioactivity release thanArmenia itself, the possibility of adverse effects is likely todeter such action. I have found no reactor expert that considersthis to be a serious threat.
One small issue deserves mention. In considerations ofreactor safety, the probability of a direct hit by a fallingaircraft is often as great as the probability of other accidents. Again this would be worse if there is no containment. Thisprobability is increased when the reactor is near an airport as itis at Yerevan. It is increased still further if, as sometimesoccurs with other landmarks, the reactor is used as a visualguidance point for incoming aircraft. ("Proceed until you see thereactor, turn left and you are in line with the runway"). Isuggest that Armenian Airlines and airport authorities be warnedagainst this common practice.
Operating staff
Although western experts did not agree with the Russianemphasis on operator error as the principal cause of the Chernobylnuclear power accident they did agree that a professional approachto operation is essential. The most important is the establishmentof what is called a "safety culture".
Every operator and staff member must think about safety in allhis actions, and must know that his supervisor will encourage andsupport him in safety actions. This is particularly hard in thosesocieties, such as the USSR, India and China, where safety has notbeen a feature of everyday life. This has been very evident in mymany visits to the USSR and was even emphasized in my travel fromArmenia to Moscow on Armenian airlines after this visit toMedzamor. Most of the passengers ignored the seat belts provided,and the belts in my seat had clearly not been used for some time. Although an announcement was made, (a small improvement from myfirst travel on Aeroflot in 1959) no effort was made by the cabinstaff to ensure that everyone sat down during take off and taxiing. Fire is a frequent initiator of accidents in all technologies, andcreates all encompassing and common mode failures. Smoking isforbidden in all western nuclear power plants and comparableindustrial facilities. Yet smoking among operators was frequentwhen western expects first visited Bulgarian and other easternEuropean plants. Therefore I, as did the European Bank team, paidespecial attention to the safety culture.
An important issue in reactor operation in a small country isa decision on the language in which everything is discussed. TheArmenians have decided that it will be Russian. This seems a wisedecision: most Armenians speak Russian, because of the twocenturies of close association, and Armenia has need of Russianconsultants especially those who designed this type of reactor.Armenia has joined the World Association of Nuclear Operators(WANO) which has an eastern European office in Moscow at theInstitute for Nuclear Power Operations. There is a specialradiotelephone link with this WANO office. This will keep Armeniain touch with the operators of other power plants of the same type. Armenia has now a Department of Regulation that is independent ofthe government, and exists to check every detail of the plantsafety.
Although the USA, in common with other G7 countries, formallyopposes the operation of these power plants, it is helping withregulation. The Office of International Programs of the US NuclearRegulatory Commission has used some funds provided by the Agencyfor International Development (AID) to help them. This hasincluded training in fire protection, and a grant of $100,000 thisyear. The Armenian Regulatory authority plan to use this money topay for consultants from Russia who understand this type of plant. Obviously this new regulatory authority has difficulty in findingcompetent staff. But they have overlooked one important source ofhelp. The Yerevan Physical Institute was founded by a greatArmenian Physicist Artem Isaacovich Alikahanian, and their verycompetent scientists may be able to help. It is noteworthy that astreet is named for the brothers Alikhanian. The other brother(who used the name Alikhanov) had helped in the design of the VVER440 reactors when head of the Institute of Experimental andTheoretical Physics (ITEP) in Moscow. Although the greatest needis for persons with operating experience, intelligent physicistshave easily learned, and led in the understanding of the importantfeatures of reactor safety.p In the USA this is evidenced, forexample, by the work of Drs Rasmussen, Koutz and Budnitz. (In myfirst lecture at the Yerevan Physical Institute in 1965, I reportedon the work of the graduate student Mr Budnitz who as Dr Budnitzwas now on the European Bank team). Armenia should not overlookthis reservoir of talent.
It is clear to me that everyone we met in Armenia from thePresident downwards is committed to the concept of a SafetyCulture. My colleagues from the European Bank commented favorablyon this to our Armenian hosts, and said that it was as good aresponse as they had in any of the eastern European reactors thatthey had visited. My personal observation, from more limitedexperience, is that the safety response of the top management issomewhat superior to that I have encountered in the USA and as goodas I encountered in my official review of the three power plants inTaiwan. The staff were friendly, frank, open and professional intheir response to detailed questions. The reactor and itsauxiliary equipment were clean, passage ways were not blocked, andI saw no sign of the tell-tale cigarette butts that can showwhether the dictates of management are in fact carried out.
The report of WANO was somewhat critical of the training ofthe operators. I believe that this has been corrected to a largeextent. There are now enough trained operators to operate fiveshifts - three shifts per day plus weekends. This should beadequate. 75% of the operators have had previous experience withreactors of this type, either in Armenia several years ago, or atNuovoVoronezh. Other younger men are being trained in thesimulator at NuovoVoronezh. This simulator is NOT a "full-scope"`simulator as is in use in western power plants but is adequate fortraining in ordinary operations and malfunctions. I also note thatIAEA will assemble a group to test the readiness of the plantoperating staff.
The first few days and weeks of any new operation arecritical. It is then that surprises may occur and criticaldecisions may need to be made. The Armenians have been lucky tosecure the services of Dr Gorbenko from the Kurchatov Institute inMoscow, who is perhaps the most experienced person in start up ofthese plants and was present at the start up of the Medzamor plantstwenty years ago. In this connection it is important to realizethat although the regulatory authority is independent of the powercompany, both are dependent on the government of Armenia. It ispossible that a problem will arise during the initial commissioningwhich will require a difficult decision: should the start up bedelayed a few months to complete some repair? or should it bestarted with less safety margin to ensure adequate electricity forthe winter? Ultimately if this eventuality occurs it must be adecision for the President. It would be well if he selects hispersonal advisers in advance.
Retrospective thoughts
It is easier to make sensible suggestions for past actionsthan for future actions. I make the comments that follow, not tocriticize those who made the best decisions that they knew how atthe time, but to lay a foundation for better decisions in thefuture.
In 1989, after the realization that the Medzamor reactors werenot as safe as desirable, there were two other good alternativesthat were not followed. Each could have provided the Armenianpeople with better protection against the energy embargo thatfollowed. Firstly, the USSR could have decided in 1989 to keep theplant running but to institute immediate safety improvements alongthe lines suggested by Hadjian's mission. Most of theseimprovements could have been made while the plant was operating orduring a normal maintenance shut down. Then by 1991, when theembargo occurred, there would have been two safe reactors inoperation. Secondly, but with far greater expense, steps couldhave been taken beginning in 1989, to commence the construction ofa replacement. This would by now be well under way.
Conclusions
I believe that the Armenian engineers and scientists have donean excellent job so far in their preparation for restart of thenuclear power plant at Medzamor. There have been several reviewsfrom international teams, and the Armenians have listened to theircomments and made appropriate changes and safety improvements. Further safety improvements are expected on a continuous basis.
No technology is ever perfect, and every natural phenomenonalso poses its hazards. The Medzamor reactor was, and will be, noexception. While many of the safety deficiencies noted above havebeen or will be eliminated, one remains. There is no containmentvessel. Moreover although there will be considerable resistance toearthquake, it will not be as great as suggested by modern designcriteria: that the reactor and crucial auxiliary equipment remainintact in the event of an earthquake that recurs every 10,000years. It will only be assured that it will be intact in a 1000year earthquake. But we must retain perspective. That thereactor is not intact does not mean an accident the size ofChernobyl. Moreover such an earthquake would be much bigger thanthe one in 1988 and would also destroy Yerevan and most of Armenia. There are other hazards of everyday life that are considerablygreater. For example many experts believe that the impact offossil fuel burning, through air pollution, on the citizens of theCommonwealth of Independent States, is every year as great as theone event of Chernobyl. The danger to the Armenian people of thepresent economic stagnation makes all these pale in comparison.
For all of these reasons therefore, I believe that theArmenian people can accept the renewed operation of the Medzamorpower plant with confidence. However I recommend that as soon asthe power plant is operative, and a normal situation of electricitysupply is once again assured, that Armenia make prudent plans forreplacement of the power plant with a more modern, safer, powerplant that meets international standards.
References
- 1 ANS, Workshop on Safety of Soviet-Designed Nuclear PowerPlants, American Nuclear Society, Chicago ILL Nov 21-22nd 1992Compilation of papers. Available from Dr. Edward Warman, Stone andWebster Corp. Boston, MA, 1992.2 World Bank (1993) reports on energy problems inLithuania, Ukraine, Bulgaria, Slovakia and Armenia, 1993.3 A. Abagyan, Extensive Private discussions with ProfessorArmen Abagyan of the Institute of Nuclear Power Operations, 1993.4 IAEA, Ranking of Safety Issues for WWER-440 model 230nuclear power plants: report of the IAEA extra budgetary program,IAEA TECDOC-640, 1992. 5 DOE, Department of Energy's Team's analyses of Sovietdesigned VVERs, US DoE Asst Sec for Nuclear Energy, September 1989 DOE/NE-0086 rev 1; Main report and analysis appendices, 1989.6 GRS, Safety Assessment of the nuclear power plantGreifswald units 1 to 4 (VVER 440-230) Gesellschaft furReactorsicherheit, GRS-78 also ISBN 3-923875-28-2, 1991.7 P. Yanev (1992) The December 7th 1988 Armenia, USSREarthquake. EQE Engineering, San Francisco, 1989.8 JCCCNRS, Record of the First Joint Meeting of the US-Russian and US-Ukraine Joint Coordinating Committees for CivilianNuclear Reactor Safety, March 30-31st 1993, at Kiev in Ukraine,USNRC SECY-93-094, 1993.9 Report from Asa Hadjian to Bechtel Corporation about thesafety of the Medzamor reactor in the event of earthquake (1989)10 R. Wilson (1995) "Nuclear Power Safety in Central andEastern Europe" Nuclear Safety in pressl11 The Framatome assessment of the Medzamor reactors (twovolumes) dated March 23rd 1993;12 The IAEA report IAEA-WWER-RD-058, dated February 25th199413 R Wilson et al (1982) "Health Effects of Fossil FuelBurning" Ballinger Cambridge, MA; see also new edition inpreparation