UPTEC ES09 003
Examensarbete 20 p Mars 2009
Modelling of the Reactor Containment Cooling in OL1 and OL2
Anna Sand
Teknisk- naturvetenskaplig fakultet UTH-enheten
Besöksadress:
Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0
Postadress:
Box 536 751 21 Uppsala
Telefon:
018 – 471 30 03
Telefax:
018 – 471 30 00
Hemsida:
http://www.teknat.uu.se/student
Abstract
Modelling of the Reactor Containment Cooling in OL1 and OL2
Anna Sand
The objective of the work presented in this report was to evaluate the possibility to make an accurate model of the cooling chains of the reactor containment in a new computation program for containment analysis.
A model of the cooling chains of the reactor containment has been built in the new computation program. A simple model of Olkiluoto 1 and 2 which had previously been built in the new computation program formed the basis of the project.
Information and technical specifications about the cooling chain and the reactor containment has been gathered, compiled and used as input to make the model accurate. The model was verified against manufacturers’ data. Finally, a benchmark test with main steam line break at full power, against the existing model has been made.
The new computation program can be used to model the cooling chain to obtain almost the same result as in the existing program. The new model takes longer time to build, though it is more detailed. On the other hand, because it is more detailed, it can describe the phenomena during a loss of coolant accident better than the existing program. There are also possibilities to make sensitivity analyses, e.g., what happens if a pump stops or a pipe ruptures.
Sponsor: Westinghouse Electric Sweden AB ISSN: 1650-8300, UPTEC ES09 003
Examinator: Ulla Tengblad
Ämnesgranskare: Michael Österlund Handledare: Lars Granström
SAMMANFATTNING
Vid de nordiska kärnkraftverken pågår just nu stora moderniseringsprojekt för att höja säkerheten och effekten. För att kunna höja effekten ytterligare måste verken optimeras.
Optimeringen kräver bättre modeller, vilket gör att kraven från kraftverksägarna ökar och därmed måste programvaran som används idag för att modellera händelseförlopp vid kärnkraftverken moderniseras alternativt bytas ut. Westinghouse Electric Sweden AB har börjat utreda möjligheten att byta ut programvaran som används idag för att göra
inneslutningsanalyser mot ett annat, modernare, beräkningsprogram.
Westinghouse har börjat verifiera det nya programmet för termohydrauliska analyser av reaktorinneslutningens kylning. Det nya beräkningsprogrammet kan modellera system mer detaljerat än det nuvarande. Det beror på att modeller i det nya programmet byggs upp komponentvis, vilket kan jämföras med det befintliga programmet som är grundat på förenklade systemmodeller.
Syftet med examensarbetet har varit att utvärdera möjligheten att utforma en korrekt modell av reaktorinneslutningens kylkedjor i det nya beräkningsprogrammet. Syftet var även att ta fram en metodik för hur en modell av kylkedjorna tas fram i programmet. Metodiken är dock enbart för internt bruk vid Westinghouse och finns således ej presenterad i denna rapport.
Begreppet reaktorinneslutningens kylkedjor innefattar de tre system som tillsammans bildar en kedja mellan havet och reaktorinneslutningen, innefattande kondensationsbassängen. Vattnet i kondensationsbassängen kyls av havet via en intermediär krets för att t.ex. reducera risken för spridning av radioaktivt material vid ett eventuellt läckage i en värmeväxlare.
En modell av reaktorinneslutningens kylkedjor har tagits fram i det nya beräkningsprogrammet för inneslutningsanalyser. En förenklad modell av Olkiluoto 1 och 2, liknande det nuvarande programmets modell, fanns sedan tidigare i det nya beräkningsprogrammet och låg till grund för arbetet. Information och tekniska specifikationer om kylkedjornas komponenter och dess funktion har samlats in, sammanställts och använts för att upprätta en korrekt modell. Fokus lades på kylkedjans logik samt värmeväxlare och pumpar. Den byggda modellen verifierades mot data från komponenternas tillverkare. Slutligen genomfördes ett jämförande test mellan den byggda modellen i det nya beräkningsprogrammet och modellen i det befintliga
programmet. Händelseförloppet som användes vid jämförelsen var ett giljotinbrott på huvudångledning vid full effekt.
Slutsatsen av det jämförande testet var att det nya beräkningsprogrammet kan användas för att modellera reaktorinneslutningens kylkedja och därmed erhålla nästan likvärdiga resultat som i det existerande programmet. Modellen som byggdes i det nya beräkningsprogrammet tar längre tid att konstruera, men är istället mer detaljerad och beskriver händelseförloppet bättre än det existerande programmet. En mer detaljerad modell gör det dock svårare att få en snabb överblick över modellen och det ökar även risken för inmatningsfel, då det är många variabler att hålla reda på. Å andra sidan kan den nya modellen vara lättare att förstå p.g.a. att det finns ett grafiskt användargränssnitt.
När modellen för reaktorinneslutningens kylkedja en gång har byggts, är det relativt enkelt att anpassa den efter nya förutsättningar, när t.ex. en pump eller värmeväxlare byts ut, eller konstruera om den för att passa en annan reaktoranläggning. Det går även att göra
känslighetsanalyser med den nya modellen. Vad händer t.ex. om en pump felfungerar, ett rörbrott inträffar eller en ventil öppnar/stänger obefogat?
Innan den nya modellen kan tas i bruk måste det dock bli accepterat att ändra
tillvägagångssättet vid beräkning av den värmeöverförda effekten från inneslutningen till havet via kylkedjorna. Strålsäkerhetsmyndigheten och kraftbolagen som äger kärnkraftverken måste övertygas om att den nya metodiken är både tillförlitlig och konservativ.
ABBREVIATIONS
BWR Boiling Water Reactor LOCA Loss of Coolant Accident LOOP Loss of Offsite Power
OL1 Olkiluoto Nuclear Power Plant, unit 1 OL2 Olkiluoto Nuclear Power Plant, unit 2 OL3 Olkiluoto Nuclear Power Plant, unit 3 PWR Pressurized Water Reactor
SCRAM Fast shut-down of the reactor (Safety Control Rod Axe Man) TVO Teollisuuden Voima Oyj
SYSTEM NUMBERS
System 314 Pressure relief system System 316 Condensation system System 321 Shut-down cooling system System 322 Containment vessel spray system System 323 Core spray system
System 327 Auxiliary feed water system System 342 Liquid waste system
System 541 Process measurements
System 652 Diesel engine auxiliary system System 712 Shut-down service water system System 721 Shut-down secondary cooling system System 741 Containment gas treatment system
CONTENTS
1 INTRODUCTION 1
1.1 Background 1
1.2 Objective 1
1.3 Method 1
2 WESTINGHOUSE ELECTRIC SWEDEN AB 2
3 OLKILUOTO NUCLEAR POWER PLANT 3
4 REACTOR CONTAINMENT 4
4.1 Design 4
4.2 Pressure-Suppression Principle 4
5 REACTOR CONTAINMENT COOLING 6
5.1 System 322 – Containment Spray System 6
5.2 System 721 – Shutdown Secondary Cooling System 7
5.3 System 712 – Shutdown Cooling Water System 8
5.4 Design Requirements 8
6 MODELLING 9
6.1 Existing Computation Program 9
6.2 New Computation Program 9
6.2.1 Introduction 9
6.2.2 Heat Exchanger Modelling 10
6.2.3 Pump Modelling 12
6.2.4 System Logics 12
7 VERIFICATION 13
7.1 Heat Exchanger Modelling 13
7.2 Flow Modelling 15
7.3 Adjoining Systems Modelling 16
7.4 System Logics 16
8 BENCHMARK TEST 18
9 CONCLUSIONS 21
10 REFERENCES 22
1 1 INTRODUCTION
1.1 BACKGROUND
The Nordic nuclear power plants are going through some major modernizations. To be able to increase the power output, an optimization of the power plant has to be made. Therefore the customers raise their demands on the codes used for modelling. To get better models of the power plant and thereby a better picture of course of events in case of an, e.g., Loss of Coolant Accident (LOCA), Westinghouse has started to investigate the possible use of another
computation program for containment analysis.
Westinghouse has started to verify this computation program for thermodynamic analysis of the reactor containment cooling, which has potential to replace the program which is used at present. The new computation program can model systems in more detail because they are modelled component by component, compared to the existing program which is based on simplified models.
1.2 OBJECTIVE
The objective of the work was to evaluate if it is possible to make an accurate model of the cooling chains of the reactor containment in the new computation program. The objective was also to define a methodology for using the new computation program to make models of the cooling chains of the reactor containment. The methodology is though only for internal use at Westinghouse and will thereby not be presented in this report.
1.3 METHOD
A model of the cooling chains of the reactor containment has been built in the new
computation program for containment analysis. A simple model of Olkiluoto 1/2 which had previously been built in the new computation program formed the basis of the project.
Information and technical specifications about the different components in the cooling chain has been gathered, compiled and used as input to make the model accurate. Relevant
information has been heat exchanger characteristics, pump data and system logic. The model was verified against manufacturers’ data. Finally, a benchmark test with main steam line break at full power, against the existing model has been made.
2
2 WESTINGHOUSE ELECTRIC SWEDEN AB
Westinghouse Electric Sweden AB was founded in 2003, but the business in Västerås was originally founded in 1969 as ASEA Atom, which latter became a part of ABB. Since 2006 Westinghouse is a part of the global Toshiba group. Westinghouse Sweden has about 900 employees, most of them situated in Västerås. [1]
There is three different business units in Westinghouse Electric Sweden AB, namely Nuclear Services, Nuclear Fuel and Nuclear Power Plants. Nuclear Services includes maintenance, repair, engineering services and methods for construction, operation and maintenance for power plants around the world. Nuclear Fuel includes fuel production for Pressurized Water Reactors (PWR), Boiling Water Reactors (BWR), Advanced Gas-cooled Reactors (AGR) and Magnox reactors around the world. Nuclear Power Plants is the business unit that includes new nuclear power plants projects. [1]
3
3 OLKILUOTO NUCLEAR POWER PLANT
Olkiluoto is situated in Eurajoki, Finland. In Olkiluoto there are two nuclear power plant units in operation, Olkiluoto 1 (OL1) and Olkiluoto 2 (OL2). A third unit, Olkiluoto 3, is under construction and there are plans on constructing a fourth unit. The power plant is owned by Teollisuuden Voima Oyj (TVO), which is a private limited company founded in 1969 to produce electricity for its shareholders at cost price. [2]
The Olkiluoto power plant unit 1 and 2 was constructed by ASEA Atom (nowadays Westinghouse Electric Sweden AB). OL1 and OL2 are two identically constructed boiling water reactors (BWR). Olkiluoto 1 has been in operation since 1978 and Olkiluoto 2 was first connected to the grid in 1980. Since then, the power plant has been running with a very high degree of reliability. The capacity factors for both OL1 and OL2 have been one of the best internationally for nearly the entire history of the power plant. Each of the units has a net electrical output of 860 MW. In 2006, the nuclear power plant in Olkiluoto produced 14 TWh, which is slightly more than 16% of all the electricity consumed in Finland that year. [2]
Olkiluoto 3 (OL3) is being built next to the existing units and according to current time
schedule the new unit will get connected to the grid during 2012. See Figure 3-1, for an image of what the Olkiluoto nuclear power plant is going to look like when OL3 has been built.
Unlike OL1 and OL2, OL3 will be a pressurized water reactor, specifically a European Pressurized-water Reactor (EPR). The net electrical output will be about 1600 MW, almost as much as the net electrical output from both of the units in operation today. [2]
There are also plans on constructing a fourth nuclear power plant unit, Olkiluoto 4. At present the environmental impact assessment and selection of the plant type is made. The Olkiluoto 4 unit could be ready to run in the end of decade 2010. [2]
Figure 3-1: Olkiluoto nuclear power plant. [2]
4 4 REACTOR CONTAINMENT
The reactor containment in a BWR is the fourth barrier, out of the total five barriers, in the defence-in-depth against dispersion of radioactive material to the surroundings. The other four barriers are the fuel matrix, the fuel cladding, the reactor tank and the reactor building. The main purpose of the reactor containment is to function as a barrier.
4.1 DESIGN
The containment is constructed like a closed cylinder with top and bottom. It is divided in three different areas; namely wetwell, upper and lower drywell (see Figure 4-1). All pipes in the containment which are pressurized from the reactor tank during operation are located in the upper drywell, except the pipes in the reactor trip system (system 354). In the wetwell, at the bottom of the containment, is the condensation pool.
In Olkiluoto 1 the condensation pool has an annular shape, which means that there is no water under the reactor tank, because the reactor has an internal primary recirculation pump. If instead the primary recirculation pumps are external, the condensation pool is often circular.
[3] The condensation pool functions as a cooler for steam blown down in the pool. The hot steam enters the pool either by the blowdown system (system 316) or during normal operation by the pressure relief pipes in system 314. When the condensation pool needs to be cooled, it is cooled by sea water via a particular cooling chain.
4.2 PRESSURE-SUPPRESSION PRINCIPLE
The containment is constructed using the pressure-suppression principle (PS) to reduce the necessary volume of the containment. The volume can be reduced because PS implies that the containment is constructed to reduce the pressure increase in the containment in case of an internal pipe failure, a so called Loss of Coolant Accident (LOCA).
In case of a pipe failure within the containment, steam will flow into the upper drywell. The pressure will then increase in the upper drywell. When the pressure in the upper drywell exceeds the pressure in the wetwell, the water column in the blowdown pipes and pressure drop in the pipes all together, steam will flow into the condensation pool and condensate. Non- condensable gas is transported to the compression area, which makes the pressure increase in the area.
The highest pressure is given when all gas has been transferred to the compression area. To decrease the pressure in the containment, the upper drywell is sprinkled with water. The sprinkled water is taken from the condensation pool, at the bottom of the containment. The pressure decreases when the sprinkle flow is able to condensate the flow from the pipe failure.
When the pressure in the upper drywell is lower than in the drywell, gas will flow back to the upper drywell via vacuum breakers. Thus the pressure in the containment levels out and thereby the pressure in wetwell decreases.
5 Figure 4-1: Reactor containment in OL1 and OL2. [4]
6
5 REACTOR CONTAINMENT COOLING
System 322, 721 and 712 together builds the chain that cools the water in the condensation pool with sea water via an intermediate loop, see Figure 5-1. When the temperature in the condensation pool has risen above 20ºC, the pumps in system 322, 721 and 712 is started manually. If the temperature would rise over 23ºC, a high temperature alarm will be initiated and all pumps will start automatically. The pumps force water through the heat exchangers in each circuit, which transfer the heat from the condensation pool water to the sea water in two steps. The pumps are stopped when the temperature in the pool has decreased to at least 20ºC.
Figure 5-1: Emergency cooling systems in OL1. [4]
5.1 SYSTEM 322 – CONTAINMENT SPRAY SYSTEM
The system 322 in Olkiluoto 1 and 2 is the containment vessel spray system and it is the system in the cooling chain that is situated nearest the condensation pool. It consists of four parallel, separate and almost identical circuits. Each circuit contain one pump, one heat
7
exchanger and a semi-circular spray pipe in the drywell and a quarter-circle formed spray pipe in the wetwell of the containment vessel. The circuits are physically separated and each circuit is placed in an area known as H bay. The H bay implies that individual faults, for instance an external pipe rupture in one circuit, will not cause failures in other circuits.
The functions of system 322 are to:
Cool the condensation pool during normal operation
Empty the water in the condensation pool via the liquid waste system, system 342
Fill up the containment with water after an accident
Supply the condensation pool with water from system 342 and to remove water from the pool to system 342 when the pool water has to be changed
Reduce the pressure inside the containment vessel in case of an accident, by condensing the steam in the drywell and the wetwell and cooling the condensation pool
Flush down condensable fission products in case of an accident inside the containment vessel
Transfer the water in combination with the pressure relief system (system 314) from the fuel handling pool to the condensation pool, when emptying the handling pool
Cool the condensation pool during and after steam relief together with isolation and shutdown of the reactor via system 314
Fill the lower part of the drywell with water from the condensation pool and cool this water by recirculation, whenever necessary
Spray the water surface in the condensation pool in combination with purging with nitrogen
Cool the reference line of the reactor water level measurement in the process measurements system, system 541
5.2 SYSTEM 721 – SHUTDOWN SECONDARY COOLING SYSTEM
System 721 is the shutdown secondary cooling system. It is the intermediate system between the condensation pool and the sea. The water in the system runs in a closed loop. The reason why there is a secondary shutdown cooling system is to avoid the escape of radioactive
material to the surroundings or inflow of salt water into the primary systems in case of leakage through one of the heat exchangers.
The system consists of two independent, almost identical, parallel circuits. Each circuit
includes two diesel-backed pumps, rated at 50% of the circuit flow individually. Every circuit supplies one heat exchanger in the shut-down cooling system (system 321), two heat
exchangers in system 322 and one of the air coolers in the containment gas treatment system
8
(system 741). Each circuit in system 721 includes two air coolers, which cool the pump motors of system 322, the core spray system (system 323) and the auxiliary feed water system
(system 327).
System 721 has the main function to supply the heat exchangers in systems 321 and 322 with cooling water, to maintain the operation of these systems. System 721 is also designed to provide cooling water to the air coolers for the pump motors in system 322, 323 and 327, the air coolers in system 741 and the heat exchangers in system 327.
5.3 SYSTEM 712 – SHUTDOWN COOLING WATER SYSTEM
System 712 is the shutdown cooling water system. It is this cooling system which is an open- circuit with sea water. The system consists of four independent and identical circuits. Each circuit is connected to a heat exchanger in system 721 and to the coolers of one diesel engine in the diesel engine auxiliary system, system 652. The function of system 712 is to supply the heat exchangers in system 721 and the diesel engine coolers in system 652 with necessary flow of cooling water.
5.4 DESIGN REQUIREMENTS
The cooling chain has mainly two design requirements that form the basis for sufficient cooling capacity of the circuits. The requirements have to be satisfied provided that the sea water temperature will not exceed 18ºC.
The first condition that has to be satisfied is that two circuits in system 322, out of the total four, must be able to keep the water temperature in the condensation pool below 95ºC
following a LOCA. On the other hand, all circuits must be available if the reactor is to stay in hot shutdown condition. Hot shutdown of the reactor implies that the reactor is running, but instead of leading the steam through the turbines, it is discharged within the containment where it is cooled by the condensation pool.
The second condition that has to be satisfied is made to guarantee complete condensation of the steam, in case of relief of steam by system 314 or steam blown down by system 316.
The condition is that the temperature in the condensation pool must not exceed the maximum temperature to respectively mass flow and density.
The result of both the design prerequisites is that the water temperature in the condensation pool can not exceed 55ºC during hot shutdown of the reactor.
9 6 MODELLING
6.1 EXISTING COMPUTATION PROGRAM
The existing computation program is build upon simplified models. The containment cooling chain 322/721/712 is modelled with the flow given as table input. It is assumed that the pump shaft output is deployed as heat in the cooling flow in system 322.
The heat exchangers are not modelled as components. The heat transfer rate is instead computed by using the heat transfer coefficient between the flow in system 322 and the sea water. The heat transfer coefficient is the actual capacity defined when the sea water
temperature is 18ºC and the condensation pool temperature is 30ºC. To credit the pump output and cooling of adjoining systems, the sea water temperature is assumed to be 18.3ºC, instead of the nominal 18ºC.
6.2 NEW COMPUTATION PROGRAM 6.2.1 Introduction
The new computation program has a graphical user interface. Models are built by connecting volumes and components with flow paths. Components can for instance be heat exchangers, coolers, heaters, pumps, valves and/or spray nozzles. Different parameters for the components that are needed to compute their behaviour, for instance hydraulic diameter, material and heat rate, are set by using built-in menus. Initial conditions and boundary conditions can also be set.
The containment cooling chain 322/721/712 is built by modelling the most important components, e.g., heat exchangers, pumps and some valves to model the system logics, and connect them with flow paths to each other and volumes. The volumes represent wetwell, drywell and the water volumes in the piping system. The built model can be seen in Figure 6-1.
The boxes with dashed lines are volumes and those with continuous lines, boxes marked with an F, are boundary conditions. Components marked with an H are heaters if they are placed in a volume and heat exchangers if they are placed on a flow path. Components marked with a V are valves. Spray nozzles are marked with an N.
10
Figure 6-1. Model of cooling chain 322/721/712 in OL1 and OL2, built in the new computation program.
6.2.2 Heat Exchanger Modelling
All necessary input data for the heat exchanger model in the new computation program are available from the heat exchanger specification. A heat exchanger model was calibrated in one operating point against manufacturer’s data, before it was implemented in the complete model for the cooling chain.
The hydraulic diameter for a heat exchanger was calculated from the dimensions on the specification. The hydraulic diameter is given by [5]
U dh 4Aflow
nce circumfere wetted
U
area flow Aflow
(eq. 1)
11
For the special case with plate heat exchangers, which is the heat exchanger type in the cooling chain 322/721/712 in OL1 and OL2, the hydraulic diameter is defined by
(eq. 2)
b
dh 2 bgapbetweenplates
The heat transfer rate in the new computation program was computed using the Nusselt number, instead of using the built-in equations. This was because the built-in equations based on heat transfer equations for tube exchangers, which differs from those for plate heat
exchangers. The Nusselt number is defined by [5]
dh
Nu
t coefficien ty
conductivi thermal
diameter hydraulic
d
t coefficien transfer
heat
h
(eq. 3)
Re = Reynold number (eq. 4)
Pr = Prandtl number
n
C m
Nu Re Pr
The coefficients C, m and n depend on the heat exchanger characteristics. m and n depends on the heat exchanger type and are given by the manufacturer, for a plate heat exchanger
m = 0.71-0.74 and n = 0.31.
The Nusselt number for each side of the heat exchanger was computed in a few steps in the new computation program as indicated in formula 5 to 13 below. First the mean temperature on one side of the plate heat exchanger was computed. The mean temperature was used to obtain data from a function, f(T), at this temperature. The function f(T) describes how the function of Prandtl number, viscosity and density depends on temperature. Finally the value from the function f(T) was multiplied with the mass flow on the corresponding side and the constant k. The constant k was defined by calculation by hand in one operating point. It depends on heat exchanger material and geometry, but also physical properties of the fluid.
The value of k is independent of changes in temperature and/or flow.
(eq. 5)
C m n
Nu Re Pr
n
m
flow h
A d C m
Nu Pr
(eq. 6)
m n
m m
flow
h m
A C d
Nu 1 Pr
(eq. 7)
n m
m m
flow
h m
A C d
Nu 1 Pr
(eq. 8)
(eq. 9)
m
mean m
T f k
Nu ( )
where
m
flow h n m m
flow h
A Nu d A
C d
k Re Pr (eq. 10)
12
m
flow h n m
flow h
h
A d A
d m k d
Pr
(eq. 11)
n m h
m k d
Pr
(eq. 12)
n m
Tmean
f 1 Pr
)
(
kinematicvis ity
density
cos
(eq. 13)
6.2.3 Pump Modelling
The flow in system 322, system 721 and system 712 is given by pumps. The flow
corresponding to a specific pressure drop is given by pump curves, which are defined for each pump type in the model. The rated shaft output is, in the model, added as heat to the water. In reality is the heat added to the water by the pump is less than the rated shaft output, thereby is the assumption conservative.
6.2.4 System Logics
The pumps in system 322 starts when the temperature in the condensation pool reaches set point level for high temperature alarm or when the SCRAM signal is activated. The
condensation pool is thus cooled by the cooling chain 322/721/712. The condensation pool water which has been cooled re-enters the condensation pool through return lines at the top of the condensation pool.
I-isolation implies that the isolation valves are closed. Isolation valves are situated
immediately inside and outside the containment wall, at the pipelines that goes through, i.e., on the steam lines and the feed water lines. In case of an I-isolation, starting of all pumps in the cooling chain is automatically initiated by actuation of the SCRAM signal.
If an I-isolation has been initiated, the function of the system is to spray the drywell and the wetwell, to reduce the pressure and temperature within the containment. In the model the pumps is started by a trip, which is set to the time when I-isolation is activated and the back-up diesel engines for supplying the pumps with electricity has had time to start due to a Loss of Offsite Power (LOOP).
Between the pump in system 322 and the spray nozzles in the reactor containment there are two isolation valves. One valve is situated on the flow path to the wetwell and one is situated on the flow path that leads to the drywell. The valves are opened by a trip.
The valve situated on the flow path to wetwell is closed manually between half an hour and an hour after the SCRAM signal. In the model the valve is closed one hour after the SCRAM signal, thereby the whole flow in system 322 is sprinkled in the drywell.
13 7 VERIFICATION
7.1 HEAT EXCHANGER MODELLING
A heat exchanger for system 322 was modelled in the new computation program according to the methodology described in section 6.2.2. The model was calibrated by comparing the result given by the new computation program with calculations according to basic heat exchanger theory using Microsoft Excel and the input data supplied by the manufacturers. The Excel spreadsheet has been reviewed, but has not yet been quality assured. After calibrating the model in the new computation program to one operating point and compared to Excel it is verified. This was done by changing temperatures and mass flows, one condition at a time.
First the temperature in the condensation pool was varied, while the temperature at the secondary inlet was held constant at 18ºC and the mass flow was 62 kg/s at the primary side and 78 kg/s at the secondary side. Figure 7-1a shows that the model agrees really well with the theory that the calculations in Excel is based on. To illustrate how well the model computed the heat transfer, see Figure 7-1b. The differences between the heat rates calculated with Excel and the model in the new computation program are minor.
0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
0 20 40 60 80 100
Te m pe ratu re in C on de n sation Pool [ºC ]
Heat Rate [kW]
Model Excel
-0,12 -0,1 -0,08 -0,06 -0,04 -0,02 0 0,02 0,04
0 20 40 60 80 100
Te m pe ratu re in C on de n sation Pool [ºC ]
Deviation from Heat Rate Calc. in Excel [%]
Figure 7-1a and b. Heat rate of the 322 heat exchanger made in the new computation program compared to Excel, while varying the condensation pool temperature.
Then the temperature at the secondary inlet was varied, while the condensation pool
temperature was held constant at 60ºC. The mass flow was set to 62 kg/s at the primary side and 78 kg/s at the secondary side. The result is presented in Figure 7-2a below. It shows that the result of the model in the new computation program corresponds well to the result given by the calculations made in Excel. The difference between the two calculations is plotted in Figure 7-2b. As shown are the discrepancy between the calculated heat rates is small also in this case.
14
6000 7000 8000 9000 10000 11000 12000 13000 14000
0 5 10 15 20 25 30 35
Te m pe ratu re at S e con dary In le t [ºC ]
Heat Rate [kW]
Model Excel
-0,12 -0,1 -0,08 -0,06 -0,04 -0,02 0 0,02 0,04
0 5 10 15 20 25 30 35
Te m pe ratu re at S e con dary Inl e t [ºC ]
Deviation from Heat Rate Calc. in Excel [%]
Figure 7-2a and b. Heat rate of the 322 heat exchanger made in the new computation program compared to Excel, while varying the temperature at the secondary inlet.
The third case used for verification of the heat exchanger model, calibrated in one operating point, was made by varying the primary mass flow from 80% to 120% of the nominal value.
The mass flow at secondary side was set to the nominal flow of 78 kg/s. The temperature in the condensation pool was set to 30.00ºC. At the secondary inlet was it set to 20.29ºC, which correspond to a sea water temperature of 18.00ºC. The result is presented in Figure 7-3a and Figure 7-3b.
1800 1900 2000 2100 2200 2300 2400 2500
70% 80% 90% 100% 110% 120% 130%
Prim ary Flow [% of Nom in al]
Heat Rate [kW]
Model Excel
-0,12 -0,10 -0,08 -0,06 -0,04 -0,02 0,00 0,02 0,04
70% 80% 90% 100% 110% 120% 130%
Prim ary Flow [% of Nom in al]
Deviation from Heat Rate Calc. in Excel [%]
Figure 7-3a and b. Heat rate of the 322 heat exchanger model made in the new computation program compared to Excel, while varying the mass flow on the primary side.
The fourth case for verification of the heat exchanger model in system 322 was made by varying the mass flow on the secondary side. The mass flow on the primary side was held to the nominal value of 62 kg/s. The temperature in the condensation pool was set to 30.00ºC and at the secondary inlet it was set to 20.29ºC, which correspond to a sea water temperature of 18.00ºC. The result is presented in Figure 7-4a and as seen is the results corresponding. The difference between the two computation programs, plotted in Figure 7-4b, shows that the difference is negligible.
15
2000 2050 2100 2150 2200 2250 2300 2350 2400
70% 80% 90% 100% 110% 120% 130%
Se con dary Flow [% of Nom inal]
Heat Rate [kW]
Model Excel
-0,12 -0,10 -0,08 -0,06 -0,04 -0,02 0,00 0,02 0,04
70% 80% 90% 100% 110% 120% 130%
S e condary Flow [% of Nom in al]
Deviation from Heat Rate Calc. in Excel [%]
Figure 7-4a and b. Heat rate of the 322 heat exchanger model made in the new computation program compared to Excel, while varying the mass flow on the secondary side.
It can be concluded that it is sufficient to calibrate the heat exchanger model in one operating point, using the methodology in section 6.2.2. The small deviations between the model in the new computation program and Excel are probably due to roundoff errors. The cases where the mass flow through the heat exchanger was varied did agree with Excel a little better, compared to the case when the temperatures were varied. Cases with varying mass flow were computed at the same temperatures as in the calibration point, thereby the only parameter used for computing the Nusselt number that was adjusted was the mass flow. The cases with varying temperatures were not made at the operating point used for calibration, instead the
condensation pool temperature in the second case was held constant at 60.00ºC, instead of the 30.00ºC used for calibration. The result is affected by the roundoff errors when defining f(T) for computing the Nusselt number, and thereby the heat transfer, for both sides of the heat exchanger.
Cases when temperature and flow was varied at the same time were also evaluated, with a resulting discrepancy within the same magnitude as the result presented above. The procedure with varying temperatures and mass flows was also made for the heat exchanger in
system 721, which gave results equivalent with the result presented in this section.
7.2 FLOW MODELLING
The flows in the cooling chain were obtained by adjusting the friction loss coefficient in the flow paths downstream the pump model. The flow was adjusted to correspond to the correct value at steady state, when the pump has passed the start up phase, see Figure 7-5. The sea temperature was held constant at 18.00ºC and the condensation pool temperature was held constant at 30.00ºC.
A pump flow peak is seen after about 3 seconds in the model of system 322. The pump flow peak creates a temporary pressure drop downstream the pump, which in turn reduces the pump flow to the nominal value. The peak is not to be expected in reality.
16
0 10 20 30 40 50 60 70 80
0 50 100 150 200
Tim e [s]
Flow [kg/s]
System 322 System 721
Figure 7-5. Flow in system 322 and 721, when sea and condensation pool temperature is held constantly at 18.00ºC respectively 30.00ºC.
7.3 ADJOINING SYSTEMS MODELLING
There are some adjoining system to system 721, for instance systems 323 and 327. The adjoining systems should have a flow of 2 kg/s when Valve 3 in the model (see Figure 6-1) is open, which means that the total flow in system 721 increases to 80 kg/s instead of 78 kg/s.
The desired flow to the adjoining systems is adjusted by adapting the friction loss coefficient in the flow paths. Figure 7-6 shows that the correct flow was obtained at steady state. The correct heat rate and function of the adjoining systems can not be modelled though, since the reactor tank model in the new computation program is not completed at the time being.
0 10 20 30 40 50 60 70 80
0 50 100 150 200
Tim e [s]
Flow [kg/s]
Adjoining Systems 721.HX P rimary Side
Figure 7-6. Mass flow to the primary side of the heat exchanger in system 721 and to the adjoining systems.
7.4 SYSTEM LOGICS
Between the pump in system 322 and the spray nozzles in the reactor tank there are two isolation valves. One valve, Valve 1, is situated on the flow path to the wetwell and one valve, Valve 2, is situated on the flow path that leads to the drywell (see Figure 6-1). The valves are modelled to open immediately when they get an opening signal. The valves are opened by a trip, which is initiated one second after the pump start. This is made in order to ensure that the
17
pump model has been initiated and been able to start building up the pressure, thereby avoiding water flowing backwards through the model of system 322.
One hour after the SCRAM signal, the valve situated on the flow path to the wetwell is closed.
The closure is initiated by a trip and the whole flow in system 322 is thereby sprinkled in the drywell. The valves positions can be seen in Figure 7-7.
0 1
1 10 100 1000 10000 100000
Tim e [s]
Valve Travel
Valve 1 Valve 2
Figure 7-7. Position of Valve 1 and Valve 2. 0 = Closed, 1 = Open.
18 8 BENCHMARK TEST
The new model of the cooling chain 322/721/712 in OL1 and OL2 was implemented in the TVO-model. A benchmark test between the new model, the TVO-model and the existing model was done. The case used for benchmarking was a full steam line break at full power.
References to the new model refer to the model built during this project, with the implemented model for cooling chain 322/721/712. TVO-model refers to the model, which already existed and formed the basis for this project, built in the new computation program. The TVO-model has the same simplified model of the cooling chain as the existing computation program.
One difference between the models is that the estimated heat transfer during the first hour in the existing model, and the TVO-model model, is larger than in the new model (see Figure 8-1). This is because the existing computation program assumes a constant heat transfer rate, which is a too simplified assumption. The heat transfer in the existing computation program only depends on sea temperature and temperature in condensation pool, where as the new computation program has a more sophisticated model as described in section 6.2.2.
The major difference though is that the flow through system 322 fluctuates in the new model, due to system 323 that has a quenching effect on the break flow. System 323 takes water from the condensation pool to the reactor tank and thereby cools the core. The flow from the
condensation pool extracted by system 323 fluctuates, because that flow is extracted from the pool until the water reaches a high level set point. When the water level in the reactor tank has decreased below a low level set point, due to the break flow, system 323 starts to extract water from the condensation pool again. Filling up the reactor tank from the low level set point to the high level set point by system 323, takes about ten minutes.
The water supplied to the reactor tank by system 323 cools the tank and thereby reduces the break flow temporarily. Thus the spray flow in the drywell is able to condensate the break flow, thus causing a pressure drop in the drywell. In the wetwell is there still a high pressure, due to non-condensable gases. This causes less pressure drop over the pump in system 322, and thus a lower head.
A lower head causes a higher flow temporarily by the pumps in system 322. An increased flow through the heat exchanger in system 322, gives a peak in heat transfer rate for the heat
exchanger. A higher flow through the heat exchanger also causes the temperature in the spray flow to increase, see Figure 8-2. These peaks are only temporary, thus the temperature in the condensation pool is not notably affected.
The flow increases temporarily during start up of the pump in system 322 and after Valve 2 is closed, see Figure 8-3. The mass flow in system 322 decreases in the new model when the temperature in the condensation pool rises, because the pump capacity is given by the volumetric flow.
The temperature in the condensation pool is about 1ºC lower using the new model, compared to the existing model, after 25 000 seconds and onwards. See Figure 8-4.
19
5000 10000 15000 20000 25000
0 20000 40000 60000 80000 100000 Time [s]
Heat Rate, 322HX [kW]
New model T VO-model Existing model
Figure 8-1. Heat rate for the heat exchanger model in system 322.
20 25 30 35 40
0 20000 40000 60000 80000 100000 Tim e [s]
Spray Temperature [ºC]
New model T VO-model Existing model
Figure 8-2. Spray temperature in system 322.
20
80 90 100 110 120 130 140 150 160 170 180
0 20000 40000 60000 80000 100000 Time [s]
322 Flow [kg/s]
New model T VO-model Existing model
Figure 8-3. Total flow in two circuits of system 322.
20 30 40 50 60 70 80 90
0 20000 40000 60000 80000 100000 Tim e [s]
Temperature in Condensation Pool [ºC]
New model T VO-model Existing model
Figure 8-4. Temperature in condensation pool.
21 9 CONCLUSIONS
The new computation program can be used to model the cooling chain 322/721/712 to obtain almost the same result as in the existing program. A model in the new computation program takes longer time to build, though it is more detailed. The more detailed and complicated model increases the risk for modelling errors, because there are lots of variables to handle and it is easy to make typing errors or making them refer to the wrong component. The more detailed model can also make it harder to get a quick overview of the model composition.
On the other hand, if one acquaints oneself with the new computation program and the specific model, the new model of the cooling chain 322/721/712 is better represented and it is easier to obtain an understanding of the system logic used for the model, because there is a graphical user interface and all the components are represented by symbols.
The model in the new computation program would be much easier and faster to understand if there were fewer variables. It would give a model which is more straightforward and reduces the risk of, e.g., typing errors. If there were built-in equations for computing the heat transfer rate for plate heat exchangers, many of the variables added especially for this model could be removed. The pump has now been modelled in a unnecessary complex way. If the built-in pump in the new computation program worked properly for our application, a few more variables could be removed.
Once the model for the cooling chain 322/721/712 in OL1 and OL2 is built, it is easy to modify when a component, e.g., a pump or heat exchanger is replaced. It can be done by modifying parameters of the existing components or by adding and/or deleting components.
The model can also be reused for another nuclear power plant by adapting the components and their logic.
One other advantage with the model in the new computation program, compared to the existing one, is that it can be used to make sensitivity analyses, e.g., what happens if a pump stops, a pipe ruptures, there is a flow obstacle or a valve does not open and/or close.
The new model for the cooling chain 322/721/712 is worth implementing. It describes the phenomena during a LOCA in more detail compared to the existing model, because fewer assumptions are being made about the heat transfer rate. It is also easier to perform sensitivity analyses, which can be a benefit if the customer demands it in the future.
Before the model built in the new computation program can be implemented though, it must be accepted to change the computation procedure of the heat transfer rate. The customers must be convinced that the model is both trustworthy and conservative.
22 10 REFERENCES
[1] Westinghouse Electric Sweden AB, Välkommen till Westinghouse i Västerås!, April 2008
[2] TVO homepage, www.tvo.fi, 2008-09-25
[3] SKI Rapport 03:02, 2003 Störningshandboken BWR
[4] TVO, Nuclear Power Plant Units Olkiluoto 1 and Olkiluoto 2, January 2008 [5] Ekroth, Granryd, Tillämpad Termodynamik, 1999
