Geothermics, Vol. 16, No. 2, pp. 197-211, 1987. Printed in Great Britain.

0375--6505/87 $3.00 + 0.130 Pergamon Journals Ltd.

(~ 1987 CNR.

DESIGN AND PERFORMANCE OF DIRECT HEAT EXCHANGE GEOTHERMAL DISTRICT HEATING

SCHEMES

R O B E R T H A R R I S O N

Energy Workshop, Department of Physical Sciences, Sunderland Polytechnic, Langham Tower, Ryhope Road, Sunderland, SR2 7EE, Tyne and Wear, U.K.

(Received November 1986; accepted for publication December 1986)

A b s t r a c t - - A variety of geothermal district heating scheme designs have been studied, differences of configuration have been identified and the design principles used to obtain maximum geothermal heat supply have been defined. The main principle is that return temperatures to the heat exchange must be as low as possible and to achieve this the network must be operated with variable temperature and flow in response to fluctuating demands. The location of back-up boilers, the type of sub-station and the inclusion of domestic water heating normally have small effects on performance. However, in some cases, water heating can have a detrimental effect.

PD P~= M s =

Mg = M n =

M~= R = E x =

N = Tgi=

N O M E N C L A T U R E

heat demand W geothermal heat transfer W smallest heat flow capacity through the heat exchanger W°C- heat flow capacity of the geothermal fluid W°C- l heat flow capacity of the network fluid W°C-1 heatflow capacity of the user circuit W°C -t ratio of smallest to the largest flow through the heat exchanger heat exchanger effectiveness number of transfer units of the heat exchanger---dimensionless geothermal well head temperature °C

Tgo = geothermal fluid return temperature. Tn; = heating network input temperature. Tno = heating network return temperature °C D = heating scheme demand coefficient W°C -l ~n = heating network "base" temperature ~u = heater base temperature °C Tu; = heater input temperature °C Tuo = heater output temperature °C S,; = network supply temperatq/'e regulation coefficient Sno = network return temperature regulation coefficient Su; = heater supply temperature regulation coefficient Suo = heater return temperature regulation coeffÉcient AT = effective temperature difference across the building fabric °C 6 T = temperature adjustment due to incidental gains °C Tx = external temperature °C T, = required internal temperature °C.

I N T R O D U C T I O N

G e o t h e r m a l d i s t r i c t h e a t i n g h a s b e c o m e a we l l e s t a b l i s h e d t e c h n o l o g y in r e c e n t y e a r s a n d ,

w h i l e t h e R e y k j a v i k d i s t r i c t h e a t i n g s y s t e m is p r o b a b l y st i l l t h e b e s t k n o w n e x a m p l e , t h e r e a r e

n o w m a n y m o r e o p e r a t i n g s c h e m e s a n d m a n y o t h e r s w h i c h a r e b e i n g d e v e l o p e d o r a r e b e i n g

s e r i o u s l y s t u d i e d . T h e m a j o r g r o w t h in a c t i v i t y h a s b e e n in F r a n c e , b u t t h e r e h a v e a l so b e e n

197

198 R o b e r t t t a r r i son

significant developments in the U.S.A. In 1983 Agence Fran~aise pour la Maitrise de l 'Energie claimed that, by the 1990s, France will have a geothermal heating capacity which is equivalent to an annual saving of 106 tonnes of oil (AFME, 1983). This represents about 300 separate schemes and, while the forecast may be optimistic in the light of recent reductions in oil prices, it is clear that a major programme is under way.

A considerable body of literature, most of it unpublished, now exists on the design and feasibility assessment of individual schemes. Also a number of modelling studies have been carried out in attempts to identify the basic economic conditions which are required for viable schemes. Some of these are important contributions, in particular, the work of Electricitd de France (EDF) (Lam6the-Parnaix et al . , 1980) in the European context and the Johns Hopkins University (Baron et al . , 1980) in the U.S. should be mentioned. Again, much of this work has not yet been published in journals; reviews have been carried out by Lockwood (1982) and by Lease (1982). The most recent modelling study has been carried out by the author and his co-workers (Harrison et al . , 1984). This has built upon and extended the work of EDF. A number of accounts which touch upon the principles of design have appeared in the open literature. In an early paper Einarsson (1973) discusses the advantages of base load operation. Olivet (1982) describes many aspects of the engineering and the design of schemes derived from consulting experience in the French geothermal industry and Anderson et al. (1979) give a similar account in the U.S. context. However, none of these gives an organised account of the factors which affect the design and performance of the diverse variety of geothermal heating schemes in the detail which is now possible. It is the aim of this paper to remedy this deficiency for one class of scheme.

The work reported here is one of the results of a recent study carried out for the E.E.C Commission. This has included the modelling work referred to above and also a survey of a large number of assessments of geothermal schemes. The survey has used information from a number of pre-feasibility assessments carried out for Geochaleur, the French public management organisation responsible for geothermal developments. A number of U.S. scheme assessments prepared as part of U.S. government programmes were also studied. In all, 45 separate scheme assessments were examined providing the wide perspective which is necessary to formulate useful systems of classification and to identify basic design principles. There are many different types of geothermal heating scheme; indeed each scheme is unique in some of its aspects. However, major patterns can be observed and systems of classification proposed. This work has been mainly concerned with the relationship between scheme design and performance and the classification system advanced here reflects this.

At the broadest level schemes can be classified on the basis of the compatibility of the temperature of the geothermal fluid and the supply temperatures of the heating systems and also the size of the heat load in relation to the thermal capacity of the wells. With fluid temperatures which are higher than the supply temperatures of the heating system all of the demands of some group of users can be met. Although, as Einarsson (1973) and others have shown, the geothermal heat delivered can be increased by meeting only the base load demands of a larger group of users. With lower fluid temperatures only part of the heat load can be met, regardless of the scheme size. Two basic types of scheme can be identified and these can be explained by reference to Figs 1 and 2. These summarise information drawn from the survey of French and U.S. schemes mentioned above.

Figure 1 shows well head fluid temperatures versus well depth. Clearly many of the U.S. schemes employ high temperature fluids from shallow wells. Figure 2 shows how the theoretical well powers (calculated assuming a return temperature of 25°C) relate to the peak heating demands of the schemes. In the U.S. cases the theoretical well powers are typically greater than the peak powers of the heat loads and thus the thermal capacities of the wells are being

Direct Heat Exchange Geothermal Heating Schemes

• Frencl~ schemes

D U.S. schemes

200 -- / O 0 * C km -I

/ []

rn • "~ [] rl D QQ

oV •

I I

Wett depth (m)

Fig. 1. Wellhead temperatures and well depths for geothermal heating schemes.

199

• French schemes

[] U.S schemes

o

I00-- D O

' ° ° [] ° Z " ":

, / I I I I IO IO0

Scheme power ( I~ )

Fig. 2. Theoretical well power and scheme peak heat load.

underutilised. In the French cases the theoretical well powers are significantly less than the peak powers of the heat loads. The typical characteristics of the U.S. schemes are high fluid temperatures available from shallow inexpensive wells and low heat load densities with high connection costs. In these cases the geothermal fluids meet all of the heating demands including the peaks. The French schemes, on the other hand, have different characteristics. The thermal gradients are lower and only moderate temperature fluids are available from deep expensive wells. The fluids may not be capable of meeting the full heating demands of the schemes and

200 Robert ttarrison

(a) Direct heat exchange

Geothermal fluid I I I I oo0 I ; ; I

II

Supply main 1

T~,

Primary heat exchanger

Return main

(b) Heat pump assisted heat exchanger

:tuid [

oop I I I

~° I t Primary heat

exchanger

-- 7--[ Condenser ]

i

Evaporator

I I Heat pump

E

nU~ ewr°r k I

Substations

User network

PS~ stations

(c) Heat exchange by heat pump only

GeotTi'er ma l [ ~ ' f luid I I i l

loop Ii ii t Heat

exchanger to protect the evaporator

Heat pump

J User network

Substations

7-.°

Fig. 3. Main types of geothermal heat supply.

developments are feasible only because large heat loads are available and connection costs are relatively low. In these conditions the most economical approach is to use the geothermal fluid to meet the "base" heating loads. The peak heat loads are supplied by fossil fuel fired back-up boilers.

In schemes where the full demands of the heat load are met by the geothermal fluid the geothermal heat supply is limited by the size of the heat load and cannot be improved by sophisticated design measures. Thus the design of this class of scheme is relatively simple. In schemes where the geothermal fluid supplies the base heating loads the process of design is more

Direct Heat Exchange Geothermal Heating Schemes 201

complex. This class of scheme tends to be closer to the margin of economic viability and it is important to obtain maximum performance in terms of the displacement of fossil fuel. This has major implications for the engineering design of the schemes. It is these aspects which are considered in this paper.

Schemes can also be classified on the basis of the methods used to extract heat from the fluid. Three groups can be identified, as illustrated in Fig. 3.

- - Schemes employing passive heat exchangers only (Fig. 3a). This is similar to conventional heat recovery.

- - Schemes where the heat exchanger is assisted by a heat pump (Fig. 3b). This is an unusual arrangement which seems to be peculiar to geothermal applications.

- - Schemes where the heat transfer is entirely effected by the heat pump (Fig. 3c). This is similar to the conventional heat pump configuration.

This paper deals only with schemes employing passive heat exchangers. Heat pump schemes will be considered in a later paper.

PRIMARY H E A T EXCHANGERS IN G E O T H E R M A L SCHEMES

Counter flow plate heat exchangers are usually used in French geothermM schemes. The basic arrangement is shown in Fig. 3a. Normally the geothermal temperature and the flow will remain fixed while the return temperatures and the flows from the heating network fluctuate as the heat demands of the users change.

The geothermal heat transfer is given by

Pg = M, Ex(Tgi- Trio). (1)

This is an important equation; it governs the design of this type of scheme. The basic aim is to design and operate the scheme so that the values of M s, Ex and Tno obtained give the highest feasible values of Pg.

For a counter flow plate heat exchanger the effectiveness is given by

[1 - exp {-N(1 - R)}] Ex = [17 -~ )~-~xpTZ--~l --" R')-}I (2)

[see for example Pitts and Sissom (1977)], where R is the flow balance across the heat exchanger and N is the number of transfer units, a dimensionless quantity related to the surface area of the heat exchanger. R varies from zero to one and as R increases E falls. R is determined by the size of the scheme in relation to the geothermal flow. N is a design variable, increasing N increases E. It has been shown, Villaume (1980), for schemes with high well costs, of the type being considered here, that minimum unit costs are obtained with high effectiveness. 5 NTU is a typical choice of heat exchanger size.

The way in which the geothermal heat transfer depends upon the network flow is important. Figure 4 shows the variation in a typical case. At low flows the network flow is the smallest flow through the heat exchanger

and

M~ = M~

(3) Pg = M, Ex(Tgi - T~o).

Therefore, the geothermal heat supply is limited by the network flow and is very sensitive to any changes in it. As the network flow rises the flow ratio "R" rises and the heat exchanger

202 Rober t Harr ison

-6 E

Perfect neat tronsfer

h 0 5

L I 2 5

Fig. 4. Typical variation of geothermal heat transfer with the flow ratio across the primary heat exchanger.

effectiveness falls, reaching a minimum when R = 1 and the network flow equals the geothermal flow. At high flows when the network flow is greater than the thermal flow

M, = M~, and (4)

P~ = MgE, ( T~,i - T,,o). Now the heat exchange is limited by the geothermal flow and it is no longer sensitive to changes in network flow. In the type of scheme which is being considered here the geothermal flow is expensive to produce and it is not sensible to operate the scheme in the region where the network flow is less than the geothermal flow. In this region the network flow is not large enough to absorb all of the heat which is available from the geothermal fluid. This is wasteful, the geothermal flow could be reduced without significantly affecting performance.

In general any conditions in which the network flow falls below the geothermal flow are detrimental to the performance of the scheme. This leads to the first basic principle of the design and operation of these schemes. For best performance the scheme must be designed and operated so that

network flow > geothermal flow.

Then equation (4) gives the geothermal heat transfer. Clearly, the geothermal heat transfer given by equation (4) is very sensitive to the network

return temperature T,,,,. In order to obtain the maximum heat transfer 7",,o must be kept as low as possible at all times. This is the second basic principle of the design and operation of these schemes.

The implications of these principles will now be considered.

LAYOUT AND REGULATION OF G E O T H E R M A L DISTRICT HEATING N E T W O R K S - - G E N E R A L PRINCIPLES

In a geothermal district heating scheme the main heat load is space heating and this may be supplied by a variety of heaters of different types with different temperature characteristics. Domestic water heating may be an additional, minor component of the heating load, and, occasionally, applications such as the heating of swimming pools may also be included. The

Direct Heat Exchange Geothermal Heating Schemes

Centro ti sed back-up bo i le r

203

eo~hermat

Primary heat exchanger

7".o = 41"C .l

f low = 273 m ~ h -I

To -- 8 5 ~

f l ow = 120 m 3 h -j

SchooLs and } commercial + / 249 dweLLings /

r,, =85"c 7-,o =74"c | J i

- 1 744 dweLti ngs T., = 5 2 " C

Tug = 41"C

f low = 360 m 3 h -~

t First branch

f Low =

1 5 3 m 3 h -

1264 dweLLings

T., = 52

7~o = 4 1

Second branch

Fig. 5. Schematic layout of the Garges Nard heating network.

network flow which is returned to the heat exchanger is a mixture of the return flows from all of these applications. In the main the return temperatures and flows at the heat exchanger depend upon

- - the arrangement of the different types of heater on the network, - - the regulation of the heaters to match fluctuating demands caused by changes in

temperature, - - the responses of the network to users turning off their heating systems. A wide range of different heating elements are in use in France. Many users employ

conventional radiators. At full load these operate with supply temperatures of 90°C and return temperatures of 70°C. There are also significant numbers of users who employ low temperature floor heaters. These operate with supply temperatures of about 55°C and return temperatures of 41°C at full load. A variety of other heaters are also encountered which have operating temperatures somewhere between these extremes. In a scheme which has both high and low temperature users connected in significant numbers on the same network an advantage can be gained by connecting them in series so that the low temperature users are supplied by the returns from the high temperature users. Figure 5 shows schematically the Garges Nard network (Geochaleur, 1982) where this approach is used. By connecting some of the low temperature users in series with the high temperature users the return temperature from the first branch is reduced from 74°C to 41°C at peak load. The result is that the overall network flows and return temperatures are both reduced. Cascading in this way can introduce problems of temperature

204 R o b e r t Harr i son

8

P ~t r.o E

v

8 T

. . . . . . . . \

\

\

AT

P~ a AT

eotherrno( power

DemGnd durotion

Fig, 6. Regulation of heater operating temperatures and the effect on geothermal heat supply.

compatibility which may require feedback and mixing. This is the case in the Garges Nord network. The supply temperatures to the low temperature users in both branches are too high. The required temperature is obtained by mixing in some of the return fluids. These are minor complications and they are more than outweighed by the improvements in performance which are obtained. The French will engineer cascade arrangements whenever possible, often going to great lengths.

Heater regulation is an important aspect of the operation of geothermal schemes. An example of the normally adopted regime is shown in Fig. 6. As the external temperatures rise, and the heat demands fall, the heater supply temperatures are reduced. If the flows through the heaters are kept constant the return temperatures also fall. Heating demands can be measured by the effective temperature difference across the building fabric AT. This is the simultaneous temperature difference adjusted to take account of incidental gains.

A T = 7",- b T - T,.

The instantaneous power demand of a building or a group of buildings is

P,i ~c ~ T .

By making some reasonable simplifying approximations, it can be shown that heaters with linear characteristics such as those shown in Fig. 6 where

T.i = T,, + S , . A T

and

T,. , = i , , +

are capable of meeting the full heat demands at all values of ATgiven the appropriate values of ~'.. S,., and S,,i. A full analysis of this can be found in Harrison and Mortimer (1985),

As will be seen later, control of the supply temperatures of the heaters is obtained either by regulating the network supply temperature as a whole or by regulating the temperatures of the fluids supplied to individual buildings from the heating substations. The way in which this is done is of no great significance for the performance of the system. However, it is important that the return temperatures of the heaters, which are the lowest temperatures on the network, should

Direct Heat Exchange Geothermal Heating Schemes 205

determine the return temperature of the network main. This requirement governs the way in which the network must be operated to respond to users shutting down their heating systems. In conventional fossil fuel fired district heating networks it is common practice to operate the network with essentially constant network flow. The users are fitted with bypass connections and when they shut down the redundant flow passes directly to the return main. This increases the return temperature and while this is of little significance in a fossil fuel fired heating network it would be detrimental to the performance of a geothermal supply. Clearly this bypassing must be avoided in heating networks which include geothermal heat exchangers. The network flow must be reduced when users shut down their heating systems so that there is no redundant fluid and no requirement for bypassing. Reductions in network flow have only a small effect on the geothermal heat transfer provided that the number of users shutting down at any time are not large enough to reduce the network flow below the geothermal flow. In some networks which include commercial and public buildings a large proportion of the users may be shut down at night, or during weekends, producing low network flows. In these cases it may be feasible to use fluid storage. This is discussed further below.

The general rules which must be observed to obtain the best performance from the heating system are that the network must be operated with

- - varying temperature in response to changing external temperatures - - varying flow in response to changing numbers of users. The characteristic behaviour of the geothermal supply is as shown in Fig. 6. Here it is assumed

that the network contains a large number of identical users which are shutting down at random so that at all times the number of users connected remains constant. Then the power demand is proportional to AT

Pd = DAT.

The geothermal heat supply is lowest at maximum heat demand and as AT falls T~o falls and P~ rises. A simple expression can be derived for P~

Pg = MgEx(Tgi- ~ ) - MgExS~oAT.

Diversity in real schemes In typical geothermal schemes, such as those shown in Figs 3 and 5, the primary heat

exchanger is centrally located and is connected to the network by a single set of supply and return mains. Groups of users all using the same heaters are connected to the network through heating substations. Within the context of the basic principles outlined above, network flows, supply temperatures and return temperatures will also depend upon many details of the layout of the scheme. As indicated above, the relative arrangements of blocks of heaters of different types on separate network branches is of prime importance in determining network temperatures and flows. The arrangements may be complex and in order to forecast the performance of any scheme the network temperatures and flows should be analysed in detail. Although this analysis can be lengthy it is based upon intrinsically simple principles of continuity and mixing. These are described in Harrison and Mortimer (1985) and will not be considered specifically here. There follows a discussion of a number of other different scheme arrangements which can affect performance to varying degrees.

Type of substation These may be mixing stations in which the users are connected directly to the network through

some arrangement of valves and tanks. Because bypassing is not allowed the return temperatures from these substations will be identical with the return temperatures of the heaters. Mixing may

206 Robert Harrison

Geothermat Loop

I I

I I

T /k,

Back-up boi ler

7-ul

Substation Group I

Substation Group 2

o

E Io0

C

g 0

50

T~o f I

& ~/[ = 0 .8

06 i 14

M~

M.

Fig. 7. Inclusion o f secondary beat exchangers at substations. (a) Schematic beat layout. (b) Effect o f f low rat io at the secondary beat exchanger on the network inlet and return temperatures.

be required to obtain compatible supply temperatures but this depends upon the location of the back-up boilers as described below. In many networks the substations house secondary heat exchangers and in these cases there is no direct connection between the heating fluids supplying the buildings and the network fluids. Secondary heat exchangers are used to reduce the pressures on the network caused by high rise buildings. However, they also have the effect of increasing the supply temperatures required from the network and the return temperatures to the network. The effect is shown in Fig. 7 where these temperatures are plotted against the flow ratio across the secondary heat exchanger. If the heat exchanger effectiveness is high (>90%) and provided that the network flow through the heat exchanger is less than 80% of the user flow, then the return temperature to the network is only about 2°C higher than the user return temperature.

Direct Heat Exchange Geothermal Heating Schemes 207

This is the only significant effect which secondary heat exchangers have upon scheme perform- ance and so long as the return temperatures are elevated by only 1 or 2°C then it is of secondary importance.

Location of back-up boilers Two basic approaches to the location of back-up boilers are encountered. When an existing district heating system is being adapted to geothermal heating, or,

alternatively, where a new network is being built and all existing boilers are being scrapped, the back-up boilers will be centralised close to the geothermal heat exchanger. The French schemes, Garges Nord and Orly are of this type (Geochaleur, 1982 and 1980). In these cases the temperature of the network supply main is regulated to meet the demands of the highest temperature users on the network, and feed back and mixing may be required to obtain temperatures which are compatible with any low temperature users. In these networks all of the heat is supplied from the central heating station and there is no way of supplementing at the substations. Then at each substation of whatever type

M,(Tn i - Tno)= Mu(T , i - Tuo).

The network temperatures and flows are essentially determined by the user temperatures and flows and because Of this these are the simplest networks to analyse. The main area of flexibility is in the choice of network supply temperature, but it seems to be normal practice to choose the lowest possible supply temperature. This reduces network losses and also reduces the possibility of oversupplying heat to the substation, ensuring elevated return temperatures are avoided. Network flows must also be carefully controlled as any redundant flow at the substations will also lead to elevated return temperatures as described above.

When large existing buildings or a collection of existing group heating schemes are connected together to form a geothermal network then the original boiler houses will be converted to heating substations and the original boilers may be retained as back-up boilers. The Fontaine- bleau scheme is of this type (Geochaleur, 1979). Thus in these schemes there is no centralised back-up, the temperatu~'e regulation is imposed at the substation and the mains temperatures follow these fluctuations in a natural way. An analysis has shown that the network return temperatures faithfully follow the user return temperatures, only being elevated by 1 or 2°C if secondary heat exchangers are used.

Now at the substations

M , ( T , i - Tno) ~ M , ( T , i - T~o).

The network flow in the geothermal loop is not uniquely determined but normally it'will be chosen so that

Mg < M n < sum of the user flows.

This has an insignificant effect on performance, and, overall, the location of back-up boilers is of minor significance for the performance of a scheme.

Provision of domestic hot water Average demands for domestic hot water are about 2001 per dwelling per day at a temperature

of about 50°C. If this is supplied by a heat exchanger, heating the water up from a cold temperature of about 10°C, then low return temperatures are possible and this makes it suitable for geothermal heating. However, it is only a small heating load (about 10% of the space heating load of the dwelling) and when supplied in parallel with space heating loads, as shown in Fig. 8,

208 Robert Harrison

Centre[ beck-up

I E Substotion

Weter ! Spoce heeting

a

°~ T n / = ' ~ ' I ~ Geothermet I-Mq-- --/------------ __ I . ' ~ ,beet

/k T Demand durations b

Fig. 8. Domestic water heating in centrally regulated networks. (a) Schematic diagram of network layout and substation arrangement. (b) Changes in network flow with changes in ATand the effect on geothermal heat supply.

the overall effect is to reduce the substation return temperature by only about 1 or 2°C below the space heating return temperatures. The effect of the inclusion of water heating on the overall performance of the scheme depends upon the type of heating network being employed. In particular it can have a detrimental effect on the performance of networks which have centralised back-up boilers and which are therefore centrally regulated as described above. In these cases the regulation regime is modified because the water heating component requires a fixed network supply temperature of about 60°C. So long as T,,i is about 60°C there are no problems, the supply temperature is regulated to follow changes in ATin the normal way and the supply temperatures to the water heaters are reduced to 60°C by feedback and mixing. The problems arise at low values of AT when network supply temperatures of less than 60°C would normally be used. In this region the network supply temperature must be maintained at 60°C to be compatible with the water heating system and the lower temperatures required for the space heating loads are obtained by feedback and mixing. The result of using feedback in relation to the majority of the heat load is to reduce the overall mains flow, and this can quickly fall below the geothermal flow. As the network flow falls below the geothermal flow the heat transfer at the

Direct Heat Exchange Geothermal Heating Schemes 209

SuppLy main

Back-up ~ w H [.~r Us!rs

t FLows in i I" /high demand ~ l

I , " periods I /

~ • Return main

Geothermat I Loop /

Dashed tine indicates fLow directions during tow demand periods

Fig. 9. Schematic arrangement of network using storage to smooth fluctuations in the number of users. Dashed line indicates flow directions during low demand periods.

primary heat exchanger is restricted as described above, and this has a detrimental effect on the geothermal heat supply at low levels of space heating demand. This will be unimportant if Tg i > 60°C because the reduction in geothermal heat supply will tend to occur at heating demands which would be oversupplied by the geothermal heat exchanger anyway. However, if Tgi < 60°C then the effect can be important and back-up heating could be required over the whole demand range. Figure 8 shows schematically the type of behaviour which occurs. The Pessac scheme is a scheme of this type and displays just this form of behaviour (CEC, 1985). The problem can be overcome by using some form of localised heating in the substation or in the dwellings to back-up the domestic water heating component .only. If this approach is feasible then disturbance of the regulation regime can be avoided. Clearly in networks which already have distributed back-up at substations no problems arise.

On the face of it domestic water heating seems to offer great improvements in the performance of geothermal schemes. However, on more careful examination the benefits appear to be mixed. As a heating load it is not usually large enough to give greatly enhanced earnings and as has been seen it may conflict with the space heating supply. Water heating demands do continue during the summer months when space heating loads are shut down, but not many networks would be kept in operation just to supply water heating. Although this is done at Melun l'Amont (BRGM, 1980). Finally, the costs of retrofitting a centralised domestic water heating supply into an existing building which is already equipped with water heaters in the individual dwellings can be high. The French do not seem to find this to be justified and it is their practice to include water heating in a scheme only if centralised supplies already exist.

Storage Various types of storage are used in geothermal schemes to smooth domestic water heating

loads so that heat exchangers supply a more or less constant load heating water from 10 to 50°C. However, storage in connection with the space heating loads is often also discussed for

schemes where significant fractions of users shut down their heating systems regularly. A possible arrangement is shown in Fig. 9. The storage tanks enable the flow through the heat

210 Robert Harrison

exchanger to be kept at a constant level when the network flow changes. In the low demand period the network flow is less than the secondary flow through the heat exchanger and the excess supply fluids are diverted to hot water storage. During high demand periods the network flow is supplemented by adding hot fluid from the storage tank. In order for the flows to balance it is necessary to store return fluids from the network when the network flow is being supplemented. These fluids maintain the flow through the heat exchanger in the low demand periods. The overall result is to increase the supply capacity of the network over and above what would be supplied without storage. This allows more users to be connected to the network. In essence the geothermal heat which would have been supplied to those users which are shut down during the night is stored and is supplied to the additional users on the network during the day. The viability of storage depends upon the numbers of users which regularly shut down their systems. Only if large numbers shut down so that the network flow falls below the geothermal flow for significant periods, thus limiting the heat transfer, is storage a useful option. If the network flow does not fall below the geothermal flow when users shut down then the geothermal heat transfer will not fall significantly in low demand periods. In this case the system is automatically transferring the geothermal heat supply that would have gone to the users which have shut down to the users which are still connected, and there is no need for expensive storage facilities.

The author knows of no scheme were storage has been installed to smooth space heating loads.

C O N C L U S I O N S

The methods of designing and operating direct heat exchange geothermal schemes to obtain opt imum performance are now reasonably well understood. The principles governing the basic arrangement as well as a number of variants can be defined.

The two basic rules are that the schemes must be designed and operated so that - - the network flow through the primary heat exchanger is always greater than the

geothermal flow; - - the network return temperature must be as low as possible at all times. Thus the network must be operated with variable temperatures and flows. Actual geothermal networks exhibit a number of differences of arrangement relating to: the nature of heating substations

- - mixing stations - - secondary heat exchangers the location of back-up boilers - - centralised back-up - - distributed back-up the provision of domestic hot water. By and large, these features are of secondary importance in that they have more significant

effects on the levels of network flow than upon return temperatures. Inclusion of secondary heat exchangers can increase substation return temperatures by 1 or 2°C and incorporation of domestic water heating can reduce return temperatures by 1 or 2°C. The inclusion of domestic water heating in centrally regulated networks can be detrimental to the overall performance of the scheme and great care should be exercised when analysing such schemes.

R E F E R E N C E S

Agence France pour la Maitrise de l'Energie (1983) La Geothermie 27, Rue Louis-Vicar-75015 Paris.

Direct Heat Exchange Geothermal Heating Schemes 211

Anderson, D. N. and Lund, J. W. (1979) Direct Utilisation of Geothermal Energy: A Technical Handbook. Geothermal Resources Council Special Report No. 7.

Baron, W., Kroll, P., Weissbrod, R. and Toth, W. J. (1980) GRITS: a computer program for the evaluation of direct use applications of geothermal energy, Johns Hopkins University, Applied Physics Lab.

BRGM (1980) Faisabilit6, Technique et Economique de la Geothermie Basse Energie en Europe, Bureau de Recherches G6ologiques et Mini6res, Orleans. 80 SGN 826 GTH.

Commission of the European Communities (E.E.C.)(1985) Geothermal Urban Heating at Pessac Demonstration Project, Leaflet No. 26.

Einarsson, S. S. (1973) Geothermal district heating. In Geothermal Energy. UNESCO, Paris. Geochaleur (1979) Ville de Fontainebleau, Project de Chauffage de Geothermique--Etude de Faisabilite. Internal

Report. Geochaleur (1980) Project de Chauffage Geothermique, Avant Project Sommaire Orley 1(94) Domaine Grazier.

Internal Report. Geochaleur (1982) Garges les Gonnesse Geothermie Avant Project Sommaire, Internal Report. Harrison, R., Lockwood, M. J. and Bryant, C. J. (1984) LEGS low enthalpy geothermal simulation computer program,

Vol. 1 User guide; Vol. 2 Description of modelling methods, Sunderland Polytechnic Energy Workshop. Harrison, R. and Mortimer, N. D. (1985) Handbook on the Economics of Low Enthalpy Geothermal Energy

Developments. Final Report on E.E.C. Contract EG/A2/O5/UKH. Sunderland Polytechnic Energy Workshop. Lamethe-Parnaix, D., Raymond, M. et Pourbaix, M. (1980) Le Chauffage Geothermique. EDF Direction Des Etudes

et Recherche, Chatou Paris. Lease, R. (1982) Integration of hydrothermal energy economics related quantitative studies, NT15 DE 83 001407. Lockwood, M. J. (1982) A review of geothermal appraisal models, Energy Workshop Report, Sunderland Polytechnic. Olivet, J. (1982) Le Chauffage Geothermique. Editions Eyrolles, Paris. Pitts, D. R. and Sissom, L. E. (1977) Theory and Problems in Heat Transfer. McGraw-Hill, New York. Villaume, M. (1980) L 'Energie Geothermique a Basse-Enthalpie pour le Chauffage des Locaux. OET.