WP EN2018-22

Costing methods for combined heat-and-

power plants fueled by zero-marginal cost
energy sources

Sarah Van Erdeweghe, Johan Van Bael and

William D‘haeseleer

TME WORKING PAPER - Energy and Environment

Last update: March 2019

An electronic version of the paper may be downloaded from the TME website:
http://www.mech.kuleuven.be/tme/research/

KULeuven Energy Institute

TME Branch
 Costing methods for combined heat-and-power plants fueled by
zero-marginal cost energy sources

Sarah Van Erdeweghea,c , Johan Van Baelb,c , William D’haeseleera,c,∗

a University of Leuven (KU Leuven), Applied Mechanics and Energy Conversion, Celestijnenlaan 300 - box 2421,
B-3001 Leuven, Belgium
b Flemish Institute for Technological Research (VITO), Boeretang 200, B-2400 Mol, Belgium
c EnergyVille, Thor Park 8310, B-3600 Genk, Belgium

Abstract

In this paper, several cost metrics for application to a combined heat-and-power plant fueled by
a zero-marginal cost energy source are studied. The mature levelized cost concepts are extended
with some novel metrics such as the levelized cost of exergy. The results are given for a geothermal
combined heat-and-power plant, connected to two different types of district heating systems and
for two scenarios for the heat and electricity prices (high and low). For a low price scenario, the
conventional costing method based on two separate prices for electrical and thermal energy is the
most appropriate. Also for a high price scenario, the conventional costing method is the most
appropriate for heat demands at low temperature. However, for higher-temperature heat demands,
the exergy costing method results in the highest revenues for the combined heat-and-power plant.
The authors recommend the use of the novel levelized cost of exergy metric as different types of
energy are priced with a single value. Depending on the amount of energy and the usefulness of the
energy type, an appropriate cost can be allocated to each product of a multi-energy system.
Keywords: renewable energy, thermoeconomic optimization, CHP, levelized cost, costing
methods, low-temperature geothermal energy

∗ Correspondingauthor
Email address: [email protected] (William D’haeseleer)

Preprint submitted to Energy Conversion and Management March 4, 2019

1 1. Introduction

2 Several cost metrics exist to indicate the economic performance of an electrical power plant. Among
3 others, the net present value (NPV), the levelized cost of electricity (LCOE) and the specific
4 investment cost (SIC) are mature concepts. The NPV gives the economic attractiveness of a project,
5 taking into account the investment costs, the operating and fuel costs, and the revenues from selling
6 electricity over the entire lifetime. The LCOE is the price for electricity that is required over the
7 entire lifetime to break even at the end of the project’s life, and the SIC is the ratio of the investment
8 costs and the electrical power output.

9 These economic metrics are also frequently used in the literature for stand-alone electrical power
10 plants fueled by geothermal energy [1–6]. Fiaschi et al. [1] performed an exergoeconomic analy-
11 sis for a geothermal organic Rankine cycle (ORC) and Kalina cycle. They considered two types
12 of geothermal energy sources; one at a low temperature of 120◦ C and one at a medium tem-
13 perature of 212◦ C. For the low-temperature source, the Kalina cycle has an electricity cost of
14 125EUR/MWh, which is 24-34% lower than for the ORC. However for the medium-temperature
15 energy source, the ORC shows better performance with an electricity cost of 88.5EUR/MWh. Aali
16 et al. [2] performed a thermoeconomic optimization of a combined flash-binary geothermal plant.
17 From the single-objective optimization towards minimal specific cost of output power, they found
18 that the cycle with R141b as working fluid has the best performance with 4.901USD/GJ. For the
19 Pareto front optimization considering the exergy efficiency and the specific cost of output power,
20 the optimal point is at 54.87% exergy efficiency and 5.068USD/GJ. Walraven et al. [3] optimized
21 the design of water- and air-cooled geothermal ORCs towards minimal LCOE. They concluded
22 that the LCOE for the water-cooled ORC is lower due to the higher net electrical power output
23 and the lower investment costs. However, if no water is available or the water price is very high
24 (> 1EU R/m3 ), the air-cooled ORC becomes better. An LCOE in the range of 60-170EUR/MWh
25 and 55-140EUR/MWh has been found for the air-cooled and the water-cooled geothermal ORC,
26 respectively, considering a decreasing brine temperature from 150◦ C to 100◦ C. Usman et al. [4]
27 compared an air-cooled and water-cooled geothermal ORC during off-design for different geograph-
28 ical locations. The optimization objective during off-design is the net electrical power output. They
29 made an economic comparison based on the SIC and the LCOE (considering the ORC only) and
30 concluded that cooling tower based systems are preferable for hot dry regions, whereas in mild

2
31 climates, air-cooling can be applied. Budisulistyo et al. [5] presented a lifetime design strategy for
32 binary geothermal power plants, taking into account the resource degradation. For the investigated
33 power plant, they found that the design for a partly degraded geothermal energy source at year 7 has
34 the highest overall NPV. Furthermore, they suggested some measures to overcome lower net power
35 output due to resource degradation. On the one hand, structural changes can be made by installing
36 a recuperator and reducing the heat transfer areas of the vaporizer and condenser at half-life. On
37 the other hand, the mass flow rates of the working fluid and the air cooling can be adjusted to keep
38 reasonable performance over the lifetime. Yilmaz [6] compared three exergoeconomic methods for
39 application to the Dora II binary geothermal power plant in Turkey. The specific exergy costing
40 method, the modified productive structure analysis and the unit system of one product method
41 have been compared for this application. The last method can only be applied for a system with
42 one product, whereas the other two methods are applicable for plants with more than one product.
43 Yilmaz found electricity production costs in the range of 35.5-43.6USD/MWh, depending on the
44 method.

45 Also different (near) zero-marginal cost renewable energy sources have been studied for electrical
46 power production [7–12]. Clauser et al. [7] compared the levelized cost of electricity for different
47 types of conventional and renewable energy sources. They concluded that geothermal power plants
48 have an LCOE which is highly competitive with conventional energy sources for regions with natural
49 steam reservoirs. In other regions, geothermal power production might be competitive if engineered
50 geothermal systems (EGS) would become a mature technology. However, more funding for EGS
51 research and for lowering the upfront risks and investment costs are required to achieve this. Tran
52 et al. [8] compared the LCOE for power plants with different conventional and renewable energy
53 sources under uncertainty of capital costs, O&M costs, system reliability and economic factors. The
54 authors concluded that the fossil fuel-based technologies have the lowest LCOE values but nuclear,
55 hydropower, biomass and geothermal are also very competitive energy sources. Furthermore, they
56 found that the addition of carbon pricing shifts the competitiveness of the different technologies,
57 with a negative impact on the fossil fuel-fired systems. Braimakis et al. [9] performed a thermoe-
58 conomic optimization for different energy source conditions (100◦ C to 300◦ C, presenting multiple
59 renewable energy sources) and used the SIC as the economic metric. They found that the SIC
60 is very variable, from 15,067EUR/kWe (source temperature of 100◦ C, 1.41kWe) to 770EUR/kWe
61 (source temperature of 300◦ C, 110.58kWe). Zhang et al. [10], Tian et al. [11] and Xi et al. [12] used

3
62 the LCOE as the economic metric in their thermoeconomic analyses. Zhang et al. [10] compared
63 different types of heat exchangers for a waste heat source of 120 − 200◦ C. They found LCOE values
64 in the range of 55 to 70USD/MWh for the optimal shell and finned tube type heat exchangers. Tian
65 et al. [11] studied different zeotropic mixtures for application in a dual-loop ORC for diesel engine
66 waste heat recovery. They concluded that an optimal LCOE of 60.3USD/MWh can be achieved for
67 a MD2M/R123 (0.35/0.65) mixture. In addition, Xi et al. [12] investigated mixtures with R245fa
68 (as flame retardant) for waste heat temperatures of 100-180◦ C. They found that the use of mixtures
69 is more economic than using a pure working fluid, mainly due to the lower evaporator investment
70 cost. The mixtures R245fa/isopentane and R245fa/pentane were recommended.

71 Application of the LCOE and SIC concepts is very straightforward for a stand-alone electrical power
72 plant since there is only one product. Similarly, for a stand-alone heating plant, the NPV, the SIC
73 and the levelized cost of heating (LCOH) can be defined. For example, in the paper of Dominković
74 et al. [13], the share of each technology in the heat production mix to a district heating (DH)
75 system is defined based on the marginal cost of each technology. In that study, the authors have
76 shown that the use of waste heat and solar energy — which have (near) zero fuel costs — leads to
77 lower marginal costs. Besides, Huculak et al. [14] and Kecebas et al. [15] have used the LCOH as
78 the economic metric for geothermal DH systems. Huculak et al. [14] studied the economic aspects
79 of six geothermal heating plants in Poland. They compared the net price of 1GJ for different fuel
80 types and concluded that the use of brown coal is the cheapest, followed by black coal. The authors
81 found that geothermal systems are less competitive but have a lower LCOH than natural gas,
82 biomass and fuel oil. The drawbacks of geothermal systems are the high risks and start-up costs.
83 Kecebas [15] performed a comprehensive analysis of the Afyon geothermal DH system in Turkey,
84 using seven wells with a reservoir temperature of ∼ 105◦ C. Kecebas concluded that geothermal
85 energy is cleaner and cheaper (0.397USD/m3 ) than fossil fuel-fired systems.

86 The allocation of costs and revenues becomes more difficult for an energy production system with
87 multiple products. While there are straightforward methods for the stand-alone heating and elec-
88 tricity plants, this is no longer the case for a multi-energy system. Some papers have used the exergy
89 costing method for the assessment of multi-energy generation system [16–19], others have used the
90 equipment cost technique (per module) [19–21]. Mehrpooya et al. [16] investigated a solar fueled
91 regenerative two-stage ORC with storage tank, and used LNG to cool the condenser. Chilled water

4
 92 is produced via the cold LNG and additional electrical power is generated by expanding the LNG
93 over a turbine. The optimal point of the Pareto front considering exergy efficiency and product
94 cost rate is at 19.59% and 3.88 106 USD/year. Akrami et al. [17] studied a geothermal system
95 providing electricity, heating, cooling and hydrogen. For a geothermal water temperature of 185◦ C
96 and 215◦ C, the total unit cost of the products are 23.18USD/GJ and 22.73USD/GJ, respectively.
97 Boyaghchi et al. [18] have also studied a geothermal (133◦ C) multi-generation system which pro-
98 vides electricity, heat for vaporizing LNG, cooling and hydrogen. They found that by optimizing
99 some operating conditions, the total avoidable exergy destruction cost rate can be improved by a
100 factor 4.9.

101 Karellas et al. [20] considered a hybrid biomass and solar trigeneration system. They concluded
102 that for providing heating, cooling and electricity to a typical Greek apartement block, the savings
103 in fuel oil and electricity consumption lead to an internal rate of return (IRR) around 12% and a
104 payback time of 7 years. Pina et al. [21] proposed a cost allocation method for the products of a
105 hybrid solar and natural gas-fired trigeneration system with thermal storage tank. They assumed
106 that all input energy flows (natural gas, solar radiation and electricity) and their respective costs
107 are known. Based on these input values, they found a cost allocation method which results in total
108 annual savings of 9,942EUR. And finally, Leiva-Illanes et al. [19] investigated a quadruple solar
109 energy system and have compared the bare equipment cost technique with the levelized exergy
110 costing method. The authors concluded that the exergy costing method is more accurate, whereas
111 the equipment cost method can be used for a quick calculation of some levelized cost indicators.
112 They calculated the levelized electricity, heating, cooling or water cost (LEC, LHC, LCC and LWC
113 respectively 1 ) for every individual module. However, no general levelized cost metric for the entire
114 multi-energy system has been proposed.

115 Also for the investigation of combined heat-and-power (CHP) plants, it seems that there is not
116 one outspoken cost metric to be used. For example, Mundada et al. [22] studied the economic
117 performance of a system consisting of off-grid solar PV panels, combined with batteries and a
118 natural gas CHP for domestic applications. They allocated all costs to the electricity generation

1 The LEC in the work of Leiva-Illanes et al. [19] is the same as the levelized cost of electricity (LCOE) in this
paper. The LHC, LCC and LWC are defined in an analogous way.

5
119 part and therefore used the LCOE as the performance indicator. Furthermore, the NPV, the IRR
120 and the annual profits have been used by Pantaleo et al. [23] and Martelli et al. [24] for a hybrid
121 solar biomass and a pure biomass CHP system, respectively. The advantage of the NPV and related
122 concepts is that the costs and revenues related to all components, fuels and products are caught
123 within one clear cost metric. From these cost metrics, it can directly be derived whether the project
124 is profitable or not. In the work of Noussan [25], different cost allocation methods are proposed
125 for natural gas-fired CHP systems. Allocations based on energy, exergy, separate generation and
126 weighted energy prices are proposed, and even special allocation methods called ”power” or ”heat”
127 bonus are discussed. However, the proposed cost allocations are based on the primary energy/fuel
128 consumption, which is less appropriate for (near) zero-marginal cost waste heat or renewable energy
129 sources like geothermal 2 .

130 In this paper, some existing and novel cost concepts for application to CHP plants fueled by a
131 (near) zero-marginal cost renewable energy source will be discussed. The existing levelized cost
132 metrics will be extended with some novel metrics, such as the levelized cost of exergy (LCOEx),
133 which properly accounts for the two products in one levelized cost metric. The different cost metrics
134 will then be applied to a CHP plant which is fueled by low-temperature geothermal energy, and
135 which delivers heat to a DH system and an ORC in parallel.

136 The novelty of this paper is that different levelized cost metrics, specific investment cost definitions
137 and costing methods are defined and compared for a CHP plant, which is fueled by a (near)
138 zero-marginal cost energy source. In contrast to the conventional fossil fuel-fired systems [25], the
139 proposed metrics are not based on the fuel consumption/price since the operating fuel costs are zero
140 or can not be clearly allocated for these types of energy sources. Furthermore, also the levelized cost
141 of exergy concept for application to a CHP plant is novel compared to the existing literature.

2 For most of the renewable energy sources, the fuel price can not be directly allocated to the energy source (which
is in contrast to e.g., gas-fired systems). For example, in case of geothermal, once the well drillings are made, the
geothermal energy source is continuously available. Note however that the pumping power costs might be considered
as operating/fuel costs in this case.

6
 Figure 1: Set-up of the parallel CHP configuration, fueled by low-temperature geothermal energy. The bold lines
indicate the path of the geothermal fluid (also called brine). EES stands for the combination of economizer, evaporator
and superheater.

142 2. Parallel geothermal combined heat-and-power plant

143 In this section, the CHP set-up, the reference parameter values and the ORC working fluid are
144 discussed.

145 2.1. Set-up

146 Fig. 1 shows the set-up of the geothermal CHP plant. The geothermal fluid (also called brine)
147 delivers heat to an ORC for electricity production and to a DH system in parallel. The brine, which
148 is indicated by the bold lines, is pumped from a production well and is pumped back in the injection
149 well after utilization. The connection to the DH system, with supply and return temperatures given

7
 Brine & wells Economic Environment Cycle DH system
◦ ◦
Tb,prod = 130 C pel = 60EU R/M W h Tenv = 10.85 C ηp = 80% Tsupply
pb,prod = 40bar del = 1.25%/year penv = 1.02bar ηg = 98% Treturn
ṁb = 150kg/s pheat = 25EU R/M W h ηm = 98% Q̇DH
Iwells = 15M EU R dr = 5% ηf = 60% psupply = 7bar
Ẇwells = 500kW L = 30years ∆Tmin = 1◦ C
N = 90% min
∆Tsup = 1◦ C

Table 1: Reference parameter values.

150 by Tsupply and Treturn , is shown in the upper part. The ORC is shown in the lower part. The
151 working fluid is first pumped from a low condenser pressure (state 1) to a higher pressure (state 2).
152 Then, the working fluid is subsequently economized (heated to a saturated liquid), evaporated (to a
153 saturated vapor) and superheated (to state 3). The three heat exchangers together are indicated by
154 EES (economizer, evaporator, superheater). Thereafter, the vapor is expanded in the turbine (to
155 state 4), generating work which is converted to electricity by a connected generator. And finally,
156 the vapor is condensed back to the saturated liquid state (state 1), which closes the cycle. This
157 cycle is continuously repeated.

158 For the heat exchangers, TEMA E shell-and-tube heat exchangers with a 30◦ tube layout are
159 considered, with the brine flowing in the tubes [26]. An axial turbine is modeled since it is common
160 in geothermal power plants and an air-cooled condenser with flat tubes and corrugated fins is used
161 [27] 3 .

162 2.2. Reference parameter values

163 Table 1 shows the reference parameter values. The brine is modeled as pure water and the values
164 related to the brine (production temperature Tb,prod & pressure pb,prod , and mass flow rate ṁb )
165 and the geothermal wells (investment costs Iwells and well pumps power Ẇwells ) are based on
166 preliminary studies for the Balmatt geological site in Belgium [28].

3 The reader is kindly referred to a previous paper [27] for a more detailed description of the different component
types and models.

8
167 Furthermore, the environment conditions (temperature and pressure Tenv and penv ) are the average
168 ambient conditions for Mol (Belgium) in 2016.

169 The parameter assumptions for the electricity and heat prices (pel and pheat ) are based on the real
170 prices in Belgium for electricity and gas (for heating) as given in the CREG report [29]. A producer
171 perspective has been assumed; so although the end-consumer prices (at the retail level) consist of
172 several contributions (energy price, grid tariffs, levies, VAT), only the bare wholesale energy price
173 is considered in this work. Other economic parameters are the yearly electricity price increase del ,
174 the discount rate dr, lifetime L and availability factor N . Lastly, 2016 is taken as the reference
175 year, also for the economic calculations.

176 The cycle parameters are the pump isentropic efficiency ηp , the motor and generator efficiencies ηm
177 & ηg , the fan efficiency ηf 4 , the minimal temperature difference over the heat exchanger ∆Tmin
min
178 and the minimum degree of superheating ∆Tsup .

5
179 Two types of DH systems are considered. A 90/60 DH system for the connection of houses with
180 conventional heating systems, and a lower-temperature 65/40 DH system for the connection of
181 houses with newer types of heating systems (e.g., floor heating) [30]. The pressure of the water
182 in the DH system pDH depends on the length and on the height differences of the DH system,
183 but a given DH system layout is assumed here. And finally, three values for the heat demand are
184 considered: Q̇DH = 5, 10 and 20M W th.

185 2.3. Working fluid

186 Isobutane is chosen as the ORC working fluid due to its good thermodynamic performance, the
187 low cost of hydrocarbons [31] and its low environmental impact [32]. Table 2 summarizes the
188 thermodynamic and environmental properties of Isobutane.

4η = 60% of Table 1 is the total fan efficiency, which includes the isentropic and the mechanical-to-electrical
f

conversion efficiency.
5 The supply and return temperatures are T ◦ ◦
supply = 90 C and Tsupply = 60 C, respectively.

9
 MW [g/mole] Tcrit [◦ C] pcrit [M P a] ODP GWP
Isobutane (R600a) 58.12 134.66 3.63 0 20

Table 2: Thermodynamic and environmental properties of Isobutane (R600a) [32]. ODP and GWP are the abbrevi-
ations for the ozone depletion potential and the global warming potential, respectively.

189 3. Thermoeconomic design optimization

190 The optimal design and the economics of the considered geothermal CHP plant are calculated based
191 on a thermoeconomic optimization model. This model is discussed first, followed by the definition of
192 the performance indicators. Lastly, the optimization results and economic performance indicators
193 are presented for the considered CHP plant.

194 3.1. Optimization procedure

195 The thermoeconomic design optimization procedure has been described in previous papers [27, 33]
196 and is used in this work. Detailed thermodynamic models for the heat transfer and pressure drop
197 calculations are implemented as well as a correlation for the turbine efficiency. Furthermore, the
198 equipment cost is calculated based on bare equipment cost functions for all components.

199 The default optimization objective is the net present value (NPV), since it accounts for the ther-
200 modynamic performance, the size and cost of the components and the time-dependency of money.
201 The variables to be optimized are the shell-and-tube heat exchangers layout (shell diameter, tube
202 diameter, baffle cut length, tube pitch and the length between the baffles), the air-cooled condenser
203 geometry (height and spacing of the fins and the number of tubes) and the operating conditions
204 (the condenser, evaporator and turbine inlet temperatures, the ORC working fluid flow rate, the
205 air velocity in the condenser and the share of the brine flow rate to the ORC branch). From the
206 optimized variables, many performance indicators, amongst which the generated electrical power
207 output and different cost metrics, can be calculated. In this paper, the focus will be on the cost
208 metrics for application to a low-temperature geothermal CHP plant connected to a DH system. But
209 the philosophy about the cost metrics can be followed for all kinds of CHP plants, and regardless
210 of the energy source.

10
211 3.2. Definition of the performance indicators

212 The NPV is the design optimization objective and is the accounted sum of costs and revenues over
213 the lifetime of the CHP plant. The NPV is defined as:
h i
X Ẇnet pel (1 + del )i + Q̇DH pheat 8760N − 0.025(IORC + IDH )
L−1
N P V = −Iwells − IORC − IDH +
i=0
(1 + dr)i
(1)
214 In the definition of the NPV, IORC and IDH are the investment costs for the ORC and for the DH
215 system heat exchanger, respectively, and Ẇnet is the net electrical power production. It is under-
216 stood that although Ẇnet and Q̇DH are power values, they are actually the respective electrical and
217 thermal energy during one timestep, being one hour. The prices are expressed per MWh. Further-
218 more, the maintenance cost can be estimated as 2.5% of the equipment investment cost according
219 to the IEA [34]. IORC , IDH and Ẇnet are dependent on the variables of the optimization proce-
220 dure. For the parameters, the reference values of Table 1 are used. The DH system temperatures
221 and heat demand are (constant) parameters to the optimization procedure. Note that a fixed heat
222 price (assuming a long-term contract) has been assumed, whereas the electricity price increases
223 with del = 1.25%/year. This is based on the European Commission report [35], where they have
224 calculated an average electricity price increase in the EU of 13% over the period 2010-2020.

225 The levelized cost of electricity and the levelized cost of heat definitions are found from the NPV
226 definition (Eq. (1)):

227 • The levelized cost of electricity, for a fixed heat price (which might be zero):
PL−1 [0.025(IORC +IDH )−Q̇DH pheat 8760N ]
Iwells + IORC + IDH + i=0 (1+dr)i
LCOE = PL−1 Ẇnet (1+del )i 8760N ; (2)
i=0 (1+dr)i

228 The LCOE is the price for electricity which is required to break even at the end of the plant’s
229 lifetime. Observe that the calculated LCOE value is the required electricity price at the first
230 year of the geothermal plant. A yearly electricity price increase of del = 1.25% is assumed,
231 similar to the NPV calculation in Eq. (1).

232 • The levelized cost of heat, for a fixed electricity price (which might be zero):
PL−1 0.025(IORC +IDH )−Ẇnet pel (1+del )i 8760N
Iwells + IORC + IDH + i=0 (1+dr)i
LCOH = PL−1 Q̇DH 8760N . (3)
i=0 (1+dr)i

11
233 The LCOH is the price for heat which is required to break even at the end of the plant’s
234 lifetime.

235 Furthermore, a new costing mechanism based on the exergy content of heat and electricity is
236
˙ DH of the DH
proposed. At this point, it is appropriate to introduce the flow exergy content Ex
237 system heat demand (Q̇DH ):

˙ DH = ṁDH (exsupply − exreturn )

Ex (4)

238 with ṁDH the DH system water mass flow rate and the specific exergy (ex) defined as:

ex = h − henv − Tenv (s − senv ) (5)

239 wherein the reference conditions are considered equal to the environment conditions (with Tenv and
240 penv as given in Table 1) in this work.

241 Analogous to the N P V which is dependent on the expected energy price, the N P Vex is defined
242 via the price of exergy. The exergy content of thermal energy and the electrical energy are priced
243 with the same price for exergy pex . Similar to the LCOE and LCOH, the levelized cost of exergy
244 (LCOEx) is defined. It is the price for (thermal or electrical) exergy which is required to break
245 even at the end of the plant’s lifetime 6 . The definitions for N P Vex and LCOEx are:

246 • The NPV, based on exergy:

N P Vex = −Iwells − IORC − IDH

h  i
X Ẇnet + Ex
L−1 ˙ DH pex (1 + del )i 8760N − 0.025(IORC + IDH ) (6)
+ ;
i=0
(1 + dr)i

247 • The levelized cost of exergy:

PL−1
Iwells + IORC + IDH + i=0 0.025(I ORC +IDH )
(1+dr)i
LCOEx = PL−1 (Ẇnet +Ex
˙ DH )(1+del )i 8760N
. (7)
i=0 (1+dr)i

248 Since the quality factor for electricity equals unity, the energy content and the exergy content are
249
˙ el ). This explains why Ẇnet is used in Eqs. (6) and (7). For heat, the quality
equal (Ẇnet = Ex
250
˙ DH is used.
factor is less than one, and the thermal exergy flow content Ex

6 As for the LCOE, the LCOEx from Eq. (7) is the required exergy price at the first year of the geothermal plant
and a yearly exergy price increase of 1.25% is assumed.

12
251 3.3. Results

252 The results are shown for a low-temperature geothermal CHP plant, connected to a 90/60 and a
253 65/40 DH system, and for three values of the heat demand. The investigation of different types of
254 DH systems is interesting because of the temperature-dependency of exergy. Multiple values for the
255 heat demand are considered since a higher heat demand means a lower electrical power generation,
256 and the effect on the project economic feasibility can be studied.

257 3.3.1. Results for two different district heating systems and for multiple heat demands

258 Table 3 shows the most important performance indicators for the design optimization of the geother-
259 mal CHP plant from Fig. 1, and for the reference parameters of Table 1. For reasons of comparison,
260 the results for the stand-alone electrical power plant (indicated by ORC ) for the same parameter
261 values are shown in the first column. Then, the results for a CHP plant connected to a 65/40 DH
262 system are shown, followed by the results for a CHP plant connected to a 90/60 DH system (which
263 all have a different optimized design). For now, only consider the values above the double bars. The
264 additional performance indicators listed underneath the double bars will be discussed in Section
265 4.2.

266 The NPV is the objective of the design optimization procedure. From the results, it follows that the
267 economic attractiveness can be increased by producing heat next to electricity for the investigated
268 parameter values, since for all cases the NPV of the CHP plants is higher than for the stand-alone
269 electrical power plant. However, the electrical power output Ẇnet decreases with the heat demand
270 and with the temperatures of the DH system. This is a direct consequence of the higher brine flow
271 rate which is needed to satisfy higher heat demands or to satisfy the same heat demand at higher
272 temperatures. As for the net electrical power output Ẇnet , the NPV is lower for the connection to
273 a higher-temperature DH system (due to the lower Ẇnet ). However, the NPV increases with the
274 heat demand as a result of the higher incomes from selling heat.

275 Also the NPV based on exergy content, as defined in Eq. (6), has been calculated. In this cal-
276 culation, the exergy price pex is assumed equal to the reference electricity price pel of Table 1.
277 N P Vex is lower than N P V because the incomes from selling heat are lower in the N P Vex cal-
278
˙ DH pex (1 + del )i ) than in the N P V calculation (∼ Q̇DH pheat ). As for the N P V ,
culation (∼ Ex
279 N P Vex increases with the heat demand due to the higher incomes from selling heat. However, in

13
 ORC 65/40 DH system 90/60 DH system
Q̇DH [MWth] 0 5 10 20 5 10 20
˙ DH [MWth]
Ex 0 0.64 1.27 2.55 0.92 1.84 3.67
Ẇnet [MWe] 3.11 2.77 2.43 1.77 2.67 2.24 1.37
N P V [MEUR] -3.74 10.33 24.53 53.00 9.82 23.52 51.06
N P Vex [MEUR] -3.74 0.02 3.91 11.76 1.98 7.85 19.72
LCOE, pheat = 0 [EUR/MWhe] 68.20 73.73 80.43 100.98 75.55 85.27 122.71
LCOH, pel = 0 [EUR/MWhth] - 47.06 22.54 10.30 46.53 22.00 9.67
LCOEx [EUR/MWhex] 68.20 59.96 52.80 41.42 56.24 46.85 33.31

LCOE [EUR/MWhe] 68.20 34.57 -8.76 -144.06 34.96 -11.65 -194.32

LCOH [EUR/MWhth] - 8.77 5.73 4.18 9.58 6.52 4.94
LCOEn [EUR/MWhen] 68.20 28.73 17.61 9.36 28.80 17.48 8.96
SICen [EUR/kWen] 8510 3301 1991 1056 3311 1982 1022
SICel [EUR/kWel] 8510 9258 10174 12976 9503 10832 15954
SICth [EUR/kWth] - 5130 2475 1149 5082 2426 1092
SICex [EUR/kWex] 8510 7529 6679 5324 7074 5952 4334

Table 3: Performance indicators (calculated based on the results of the design optimization procedure) for the parallel
geothermal CHP plant, for the connection to a 65/40 and a 90/60 DH system and for three different heat demands.
The results for the stand-alone electrical power plant (ORC) are given for comparison. To recall, the assumed price
for electricity in Table 1 was equal to pel = 60EU R/M W h, while the price for heat was pheat = 25EU R/M W h
(relevant for the NPV computation and to be compared with the levelized costs). LCOE and LCOH account for
the reference price for heat and electricity, respectively, whereas for LCOE, pheat = 0 the heat price is zero and for
LCOH, pel = 0 the electricity price is zero.

14
280 contrast to the N P V and Ẇnet , N P Vex is higher for the connection to a higher-temperature DH
281 system due to the higher exergy content of heat, so higher incomes from selling heat exergy are
282 generated.

283 Fig. 2 shows the LCOE as a function of the heat price. The lines for the LCOE (so, the electricity
284 price for which N P V = 0) of the CHP plant are given by the negatively-sloped straight lines. All
285 points above the given negatively-sloped lines are economically feasible, with a combination of pel
286 and pheat for which the CHP plant is profitable (with N P V > 0). The black full horizontal line
287 gives the LCOE for the stand-alone electrical power plant. The results for the connection to a 65/40
288 DH system are shown in blue, for three values of the heat demand: 5MWth (dotted line), 10MWth
289 (dash-dotted line) and 20MWth (dashed line). The intersection point of the blue lines with the
290 black line indicates the heat price for which the LCOE for the CHP plant equals the LCOE of the
291 stand-alone electrical power plant. For heat prices above this value, the CHP plant is economically
292 more attractive than the stand-alone electrical power plant. Note that the heat prices for which the
293 CHP plant becomes more attractive are very low: pheat = 2.39, 2.92 and 3.16EUR/MWh for heat
294 demands of 5, 10 and 20MWth, respectively. So, in almost all cases, a CHP plant is economically
295 more attractive than a stand-alone power plant for the investigated low values of the electricity
296 price (which are common values for Belgium).

297 The green negatively-sloped straight lines indicate the LCOE for the connection to a 90/60 DH
298 system. In comparison to the 65/40 DH system (blue lines), heat at higher temperatures is required.
299 From Table 3, it follows that the electrical power production is lower. As a result, for the same
300 heat demand, higher electricity prices are needed for low values of pheat . For higher values of pheat ,
301 the electricity prices are almost the same as for the 65/40 DH system. Actually, for high values of
302 pheat , slightly lower electricity prices are needed for the connection to a 90/60 DH system than for
303 a 65/40 DH system. For the connection to a 90/60 DH system, the electrical power output is lower
304 but also the investments in the ORC (IORC ) are lower which explains this trend. Furthermore, the
305 intersection point with the stand-alone electrical power plant line is at slightly higher values for
306 pheat compared to the 65/40 DH system connection. The values of the intersection points are at
307 pheat = 3.41, 3.94 and 4.15EUR/MWh for heat demands of 5, 10 and 20MWth, respectively.

308 The intersection with the ordinate axis gives the LCOE for pheat = 0. These values are also given
309 in Table 3. LCOE, pheat = 0 increases for a higher heat demand and for higher temperature levels

15
Figure 2: Levelized cost of electricity (LCOE) as a function of the heat price, for the stand-alone electrical power
plant (black), for the CHP plant connected to a 65/40 DH system (blue) and for the 90/60 DH system connection
(green). The dotted, dash-dotted and dashed lines indicate a heat demand of 5, 10 and 20MWth, respectively. The
red horizontal lines present the levelized cost of exergy (LCOEx) for the CHP plant connected to a 65/40 DH system.
In gray, the lines for the 90/60 DH system are shown. LCOE = LCOEx for a stand-alone electrical power plant
(full black line).

16
310 of the heat demand, which is logical. Similarly, the LCOH for pel = 0 is given on the abscissa axis.
311 As expected, LCOH, pel = 0 is higher for a lower heat demand. LCOH, pel = 0 is higher for lower
312 temperatures of the DH system because the electrical power output is higher, and the investments
313 in the ORC have to be covered by the revenues from selling heat. Also these values are shown in
314 Table 3.

315 Finally, the new economic metric — the levelized cost of exergy (LCOEx) — is also shown in Fig.
316 2 for the stand-alone electrical power plant (in black) and for the CHP plants with different heat
317 demands (in red for the 65/40 DH system connection and in gray for the connection to a 90/60
318 DH system). Note that the lines for the LCOEx do not depend on the heat price, since the same
319 price is assumed for one MWh of electricity as for one MWh of thermal exergy. The LCOEx is
320 defined as the price per MWh of (thermal or electrical) exergy which is required over the lifetime
321 of the project, to break even at the end of its lifetime. This metric is based on the exergy concept,
322 and makes no distinction between exergy from electrical and thermal energy. For the stand-alone
323 electrical power plant, LCOEx = LCOE. For the CHP plants, exergy is delivered by thermal or
324 electrical energy and the LCOEx decreases with the heat demand, as can also be seen in Table 3. A
325 higher heat demand means that the total exergy production is higher, so a lower price is sufficient
326 to break even at the end of the CHP plant’s lifetime. Furthermore, the LCOEx values are lower for
327 a higher-temperature DH system due to the higher exergy content of the same amount of thermal
328 energy.

329 3.3.2. Note on the temperature-dependency of exergy

330 From Table 3, it was already clear that the exergy content of the same amount of heat asked by
331 a 90/60 DH system is higher than for a 65/40 DH system. This is further explained using Fig. 3.
332

333 The specific exergy is a state property which is defined with respect to a reference state. For this
334 reference state, the average ambient conditions for Belgium in 2016 are considered. The specific
335 exergy is the amount of work that one kg of fluid (at a certain temperature T and pressure p)
336 could maximally deliver with respect to the environment at Tenv and penv (values of Table 1).
337 From Fig. 3, it follows that ex increases exponentially with the temperature. This explains why
338 the specific thermal exergy of the 90/60 DH system (ex = 23.11kJ/kg) is higher than that of the

17
 Figure 3: Specific exergy as a function of temperature, for water at pDH = 7bar and with respect to the average
ambient conditions (as given in Table 1).

339 lower-temperature 65/40 DH system (ex = 13.32kJ/kg). This is also the reason why the LCOEx
340 for the CHP plant connected to a 90/60 DH system (gray lines in Fig. 2) is lower than for the CHP
341 plant connected to a 65/40 DH system.

342 3.3.3. Note on the optimization objective

343 The result of the optimization procedure is the optimal CHP plant design which corresponds to
344 the highest NPV. However, one could wonder why the N P Vex has not been considered as the
345 optimization objective. If the N P Vex would have been the optimization objective, the results
346 (optimal CHP plant design and performance indicators, e.g., Ẇnet ) would be the same. This is
347 because (the optimal values of) the variables do not depend on the electricity or heat prices, since
348 they are (fixed) parameters to the optimization procedure.

349 4. Discussion

350 In this section, the metrics for the economic characterization of a CHP plant will be extended,
351 with each metric having its own viewpoint to the same CHP plant. Furthermore, different costing

18
352 methods will be defined and applied to two economic scenarios. Finally, some general guidelines
353 will be presented.

354 4.1. Definition of some more performance indicators

355 Some additional performance indicators are defined:

356 • The levelized cost of energy:

PL−1
Iwells + IORC + IDH + i=0 0.025(I ORC +IDH )
(1+dr)i
LCOEn = PL−1 (Ẇnet (1+del )i +Q̇DH )8760N ; (8)
i=0 (1+dr)i

357 The LCOEn is the price for (thermal or electrical) energy which is required to break even at
358 the end of the plant’s lifetime.

359 • The specific investment cost, based on total energetic power:

Iwells + IORC + IDH

SICen = ; (9)
Ẇnet + Q̇DH

360 • The specific investment cost, based on electrical power:

Iwells + IORC + IDH

SICel = ; (10)
Ẇnet

361 • The specific investment cost, based on heat (thermal power):

Iwells + IORC + IDH

SICth = ; (11)
Q̇DH

362 • The specific investment cost, based on total exergetic power:

Iwells + IORC + IDH

SICex = . (12)
˙ DH
Ẇnet + Ex
363 The specific investment cost gives the investment costs per kW of representative energy or
364 exergy which is produced. The representative energy/exergy form depends on the application.

365 4.2. Costing methods

366 From the levelized cost concepts, the LCOE is the most well-known concept. Similarly, the SIC
367 based on electricity is the most well-known. Both of them are mature concepts for electrical power

19
368 plants. However, when dealing with heat and electricity, the energy producer can choose how to
369 allocate costs. If all costs are allocated to the electricity production, the LCOE and SICen can still
370 be used. Alternatively, the costs can also be allocated to the heat production alone (LCOH, SICth )
371 or to electricity and heat production together (LCOEn, SICen ). However, the exergy concept (with
372 the LCOEx) is more appropriate 7 for a CHP plant, since there is no distinction between electricity
373 and heat. Furthermore, the amount of energy and the usefulness of heat at different temperatures
374 are taken into account.

375 The different costing methods are 8 :

376 1. Electricity costing method, either allocating all costs to the electricity production (pheat = 0)
377 or for a fixed heat price (pheat 6= 0);

378 2. Heat costing method, either allocating all costs to the heat production (pel = 0) or for a fixed
379 electricity price (pel 6= 0);

380 3. Energy costing method, using the same price for heat and electricity;

381 4. Exergy costing method, using the same price for the exergy content of heat and electricity.

382 The additional performance indicators are added to Table 3, below the double bars. Based on Table
383 3, all levelized cost concepts will be discussed for application to a geothermal CHP plant.

384 For the LCOE, pheat = 0, all costs are allocated to the electricity production. In comparison to the
385 stand-alone electrical power plant, the LCOE is higher due to the higher investment costs and the
386 LCOE is higher for the connection to a higher-temperature DH system, which is expected.

387 For the LCOH, pel = 0, all costs are allocated to the heat production. The LCOH does not exist
388 for a stand-alone electrical power plant. As is logical, the LCOH decreases with the heat demand,

7 With the term appropriate, the authors mean that the respective cost metric is the most suited for the investigated
case. The LCOEx is the appropriate levelized cost metric for CHP plants, since it accounts for the two products and
for the usefulness of every product. So it makes more sense to use the LCOEx instead of the LCOE, the LCOH or
the LCOEn for this purpose.
8 Note that the terms energy and power are used indiscriminately throughout the paper. Over the assumed

timestep of one hour, 1MWh of thermal or electrical energy is the result of a thermal or electrical power output of
1MW over a period of one hour. Prices are expressed in EUR/MWh, so the corresponding energy unit should be
used for cost calculations.

20
389 and is higher for the 65/40 DH system connection due to the higher electrical power production (so
390 higher ORC investment costs).

391 The LCOE is the levelized cost of electricity, assuming a fixed heat price of pheat for selling heat.
392 Similarly the LCOH is the levelized cost of heat, assuming a reference electricity price of pel for
393 selling electricity. The LCOE and LCOH decrease with the heat demand, since the incomes from
394 selling heat are higher. The LCOE can become negative, at sufficiently high heat prices, when no
395 additional incomes from selling electricity are required to have N P V ≥ 0. Note that the LCOE
396 decreases faster with the heat demand for the 90/60 DH system due to the lower electrical power
397 production Ẇnet . The LCOH is higher for the 90/60 DH system due to the lower revenues from
398 selling electricity (because of the lower value of Ẇnet ).

399 The levelized cost of energy (LCOEn) considers the same price for both, thermal and electrical
400 energy which, in this case, is taken as the weighted average of pel and pheat . As can be expected,
401 LCOEn is lower for higher heat demands since more energy is produced by the CHP. For low heat
402 demands, the investment costs are high compared to the energy production and the LCOEn is
403 higher for the 90/60 DH system. However for higher heat demands, the electricity production for
404 the 90/60 DH system case is lower (so are the ORC investment costs) and the LCOEn for the
405 90/60 DH system connection is lower than for the 65/40 DH system connection (where the ORC
406 investments are still significant).

407 Finally, the LCOEx concept considers thermal and electrical energy but also accounts for the
408 temperature levels of the thermal energy. The philosophy behind this concept is that heat at higher
409 temperatures, which is more useful, should have a higher price than heat at lower temperatures.
410 The LCOEx decreases with the heat demand and decreases faster when heat at higher temperatures
411 (and thus a higher exergy content) is required.

412 The SICen , SICel , SICth and SICex follow the same trends as the LCOEn, LCOE, pheat = 0,
413 LCOH, pel = 0 and LCOEx, respectively, and are also shown in Table 3.

414 4.3. Application to two different economic scenarios

415 The different costing methods will now be discussed for two economic scenarios. The first scenario
416 (REF ) is for the reference values of pel = 60EU R/M W h and pheat = 25EU R/M W h from Table

21
417 1. This scenario is suggested by the economic conditions for the wholesale prices in Belgium
418 (without some kind of support scheme) [29]. For the second scenario, pel = 250EU R/M W h and
419 pheat = 50EU R/M W h are considered. This scenario is inspired by the retail prices in Germany,
420 including feed-in tariffs (a kind of support scheme, and is indicated by SUP ) [36]. For all cases,
421 a price of exergy equal to the electricity price pex = pel is assumed and an energy price pen =
422 xpel + (1 − x)pheat , with the fraction x being the ratio of the electrical energy to the total energy
423 produced by the CHP.

424 Fig. 4 shows the actual annualized hourly revenues for every costing method. The hourly revenues
425 are calculated based on the actual prices for heat and electricity in each scenario, which are different
426 from the required prices to break even (so for which N P V = 0). Positive revenues indicate a
427 profitable project (N P V > 0) whereas negative values indicate an unfeasible project (N P V < 0).
428 Note that the CHP plant has been designed for the parameter values of Table 1.

429 The color bars represent the revenues for each costing method for the CHP plant connected to a
430 65/40 DH system and for different values of the heat demand. The first gray bar indicates the
431 stand-alone electrical power plant for comparison. The black bars without fill represent the results
432 for the CHP plant connected to a 90/60 DH system. The dashed and full lines show the maximal
433 revenues for the actual electricity and heat prices for the 65/40 DH system and the 90/60 DH
434 system connection, respectively.

435 First, consider Fig. 4a, which gives the results for typical Belgian wholesale conditions. The
436 first bar once again shows that the stand-alone electrical power plant is not economically feasible
437 for the conditions of Table 1 (revenues < 0). The other bars indicate the revenues of the CHP
438 plant based on the actual occurring prices (pel = 60EU R/M W h and pheat = 25EU R/M W h) of
439 blindly applying the costing methods. Since the actual prices differ from the required value of the
440 respective levelized cost indicator, the revenues are different from zero. For the different costing
441 methods applied to the CHP plant holds:

442 • The blind application of the electricity costing method with pheat = 0 (blue) results in
443 an economically unfeasible project. Only revenues from selling electricity are generated.
444 However, the actual price is lower than the price which is required by the cost metric:
445 pel < LCOE, pheat = 0;

22
 (a) REF: pel = 60 & pheat = 25EU R/M W h [29]. (b) SUP: pel = 250 & pheat = 50EU R/M W h [36].

Figure 4: Real hourly revenues of blindly applying the different costing methods to customers, for two different
scenarios for the electricity and heat prices. Left: scenario with low electricity and heat prices (REF ). Right:
scenario with high prices for electricity and heat (SUP ).

446 • Also, the application of the heat costing method with pel = 0 (green) results in an unfeasible
447 project for a low heat demand of Q̇DH = 5M W th. However, for the heat demands of 10 and
448 20MWth, the project is economically feasible;

449 • The exergy costing method (red) always results in a profitable project. The revenues are
450 higher for higher heat demands and for higher temperatures of the required heat (due to the
451 higher amount of exergy);

452 • The energy costing method with the weighted average price for thermal and electrical energy
453 (purple) is also always attractive, and results in higher revenues for higher heat demands and
454 lower temperatures of the required heat. The revenues are slightly lower than the maximal
455 revenues 9 , which are indicated by the dashed lines for the 65/40 DH system and by the full
456 lines for the 90/60 DH system;

457 • The electricity costing method for a fixed heat price pheat = 25EU R/M W h (yellow) and the
458 heat costing method for a fixed electricity price pel = 60EU R/M W h (cyan) correspond to
459 the maximal revenues which are possible, given the actual heat and electricity prices.

9 The energy costing method is actually the same, however, without accounting for the electricity price increase
over the years but just taking the weighted average of the prices for heat and electricity.

23
460 Note that for the considered actual electricity, heat and exergy prices, none of the costing methods
461 can blindly be applied. For all of the methods, the revenues are lower than the potential maximal
462 revenues, which are the result of the conventional costing mechanism (sell electrical and thermal
463 energy at pel and pheat , respectively). However, the application of all costing methods, except the
464 electricity costing method with pheat = 0 (blue) and the heat costing method with pel = 0 (green)
465 at low heat demands, still result in an economically attractive CHP plant (N P V > 0). The energy
466 costing method with the weighted average price for thermal and electrical energy method is closest
467 to the revenues from the conventional costing method, which nearly equals the costing methods of
468 the last two bars.

469 Fig. 4b shows the results for higher electricity and heat prices, which are typical for the situation
470 with support scheme. The first bar shows the revenues for the stand-alone electrical power plant
10
471 given the actual electricity price. For this high electricity price , the stand-alone electrical power
472 plant is economically feasible. However, the revenues and the economic feasibility of the geothermal
473 project can still be improved by providing heat next to electricity. In this case, the application of
474 all costing methods to the CHP plant results in a profitable project (however not all methods are
475 equally in demand). The trends are similar to Fig. 4a, but note the different ordinate scale. The last
476 two bars, for the electricity costing method for a fixed heat price pheat = 50EU R/M W h (yellow)
477 and the heat costing method for a fixed electricity price pel = 250EU R/M W h (cyan), reflect
478 the conventional costing method. This is the best costing method for the connection to a low-
479 temperature DH system. However, for the connection to a higher-temperature 90/60 DH system,
480 the LCOEx concept results in the highest revenues. This is mainly due to the high electricity price
481 compared to the heat price and the pex = pel consideration, and due to the higher exergy content
11
482 of higher-temperature heat. Also due to the larger relative significance of pel compared to pheat ,
483 the electricity costing method with pheat = 0 (blue) has relatively better performance compared to
484 the heat costing method with pel = 0 (green), especially at low heat demands.

485 Whether or not the costing method results in an economically feasible project (with N P V > 0),
486 depends on the actual prices for electricity and heat.

10 Note the very high electricity price (thanks to the feed-in tariffs) compared to the case without feed-in tariffs in
Fig. 4a.
11 Note that the electricity-to-heat price ratio is 2.4:1 for Fig. 4a whereas it is 5:1 for Fig. 4b.

24
 Figure 5: LCOE, pheat = 0, LCOH, pel = 0, LCOEn and LCOEx of the parallel geothermal CHP plant for the
connection to a 65/40 DH system (colored bars) and for the connection to a 90/60 DH system (black bars without
fill). For comparison reasons, the LCOE of a stand-alone electrical power plant is given by the gray bar. The
dash-dotted and the dotted lines give the actual prices for electricity and heat for the reference case (based on the
Belgian wholesale scenario with values from Table 1, indicated by REF ) and for the scenario with support scheme
(indicated by SUP ), respectively.

487 4.4. General guidelines

488 Fig. 5 shows the different levelized cost metrics for the geothermal CHP plant, for the connection
489 to a 65/40 DH system (colored bars) and a 90/60 DH system (black boxes without fill) and for
490 three values of the heat demand.

491 In general, the costing methods are profitable (N P V > 0) in the following cases:

492 • Electricity costing method with pheat = 0 for LCOE < pel ;

493 • Heat costing method with pel = 0 for LCOH < pheat ;

494 • Energy costing method for LCOEn < xpel + (1 − x)pheat , with x the share of electricity in
495 the total produced (thermal and electrical) energy;

496 • Exergy costing for LCOEx < pex . In this paper, the price for exergy pex is chosen equal to
497 the electricity price pel , but this price may be different.

25
498 For illustration, the actual prices for heat and electricity which are considered in this study, are
499 also given in Fig. 5. The dash-dotted lines present the reference prices (indicated by REF ) and the
500 dotted lines indicate the prices for the scenario with support scheme (indicated by SUP ).

501 The energy producer does not want to allocate all costs to electricity and sell heat for free. This
502 leads to a higher LCOE than for a stand-alone electrical power plant. Even more, the electricity
503 price in the reference case is lower than LCOE, pheat = 0, so this results in a negative economic
504 balance. Also the opposite is true. The producer generally does not want to sell the electricity for
505 free (since the ORC part needs the biggest investments) and generate incomes from selling heat
506 only. For the REF case, LCOH, pel = 0 > pREF
heat and this costing method does not result in an

507 economically attractive project, however for the SUP case, the heat costing method with pel = 0
508 results in an economically feasible project since LCOH, pel = 0 < pSU P
heat . It would be a lot more

509 beneficial, however, to invest in a heat exchanger for heat generation only. The LCOH would range
510 from 6.18EUR/MWh for a 65/40 DH system and Q̇DH = 20M W th to 24.08EUR/MWh for a
511 90/60 DH system and Q̇DH = 5M W th. By doing so, the LCOH would be lower than the heat price
512 in both cases. Furthermore, the producer would like to use the energy costing method. Since the
513 electricity and heat prices are generally higher than the required LCOEn-value, this price allocation
514 method results in an economically attractive project. And finally, the exergy costing method is also
515 attractive, and even more for high heat demands and for a heat demand at high temperatures.
516 Since the heat is appropriately valued based on its temperature level, this costing method is closest
517 related to thermoeconomics.

518 5. Conclusions

519 In this paper, the design optimization of a geothermal combined heat-and-power (CHP) plant
520 towards maximal net present value (NPV) has been investigated. This NPV calculation is based
521 on parameter assumptions for the electricity and heat prices. However, once the CHP plant is
522 installed, the price allocation to heat and electricity can be done according to certain costing
523 methods. Therefore, different levelized cost metrics and specific investment cost definitions have
524 been defined. Also different costing methods have been discussed and applied to a CHP plant fueled
525 by low-temperature geothermal energy, which is a (near) zero-marginal cost energy source.

26
526 A general rule states that the application of a certain costing method to the customers results
527 in an economically profitable project (N P V > 0) if the actual price is higher than the corre-
528 sponding levelized cost metric. The results are given for the reference scenario (REF ) with actual
529 prices for heat and electricity of 25 and 60EUR/MWh, respectively, and for a scenario with high
530 feed-in tariffs (SUP ) and corresponding heat and electricity prices of 50 and 250EUR/MWh. For
531 the REF scenario, the application of the conventional costing mechanism (with fixed prices for
532 heat and electricity) results in the highest revenues. However for the SUP scenario, the conven-
533 tional costing mechanism is the most convenient for the connection to a low-temperature 65/40
534 DH system, whereas the exergy costing method results in higher revenues for the connection to a
535 higher-temperature 90/60 DH system. The exergy price is assumed equal to the electricity price
536 in this work, but this value might be different. In general, the exergy costing method results in
537 higher revenues in case of higher heat demands, higher temperatures of the heat demand and a
538 larger difference between the exergy price and the heat price.

539 The authors recommend the use of the novel levelized cost of exergy (LCOEx) metric since it is
540 closest related to thermoeconomics and different types of energy products can be represented with
541 a single value. Depending on the amount of energy and the usefulness of the energy type (e.g., for
542 heat depending on the temperature level), an appropriate cost can be allocated to each product of
543 a multi-energy system.

544 Acknowledgments

545 This project receives the support of the VITO PhD grant number 1510829. The first author would
546 like to thank dr. Ben Laenen and the VITO management for making this project possible.

27
547 Nomenclature

548 Abbreviations

symbol description
CHP combined heat-and-power
DH district heating
EGS engineered geothermal system
EES economizer, evaporator, superheater
GWP global warming potential
549 LNG liquefied natural gas
ODP ozone depletion potential
O&M operation and maintenance
ORC organic Rankine cycle
PV photovoltaic
REF reference scenario
SUP scenario with support scheme

28
550 Symbols

symbol description
del [%/year] electricity price increase
dr [%] discount rate
Ė [MWth] flow exergy
ex [kJ/kg] specific flow exergy
h [kJ/kg] specific enthalpy
I [MEUR] investment cost
IRR [%] internal rate of return
L [year] lifetime
LCOE [EUR/MWh] levelized cost of electricity
LCOEn [EUR/MWh] levelized cost of energy
LCOEx [EUR/MWh] levelized cost of exergy
LCOH [EUR/MWh] levelized cost of heating
ṁ [kg/s] mass flow rate
551

M W [g/mole] molecular weight

N P V [MEUR] net present value
N [%] availability factor
pel [EUR/MWh] electricity price
pheat [EUR/MWh] heat price
p [bar] pressure
Q̇ [MW] heat
SIC [EUR/kW] specific investment cost
s [kJ/kgK] specific entropy
T [◦ C] temperature
v [m/s] velocity
Ẇ [MW] electrical power
x [-] ratio of electrical to total energy
η [%] efficiency

29
552 Subscripts & superscripts

symbol description
b brine
crit critical point
el electricity
en energy
env environment
ex exergy
f fan of the condenser
g generator
inj injection state
553 m motor
min minimum
net net value
p pump
prod production state
return return of DH system
sup degree of superheating
supply supply of DH system
th thermal
wf working fluid
wells well drillings

554 References

555 [1] D. Fiaschi, G. Manfrida, E. Rogai, L. Talluri, Exergoeconomic analysis and comparison be-
556 tween ORC and Kalina cycles to exploit low and medium-high temperature heat from two
557 different geothermal sites, Energy Conversion and Management 154 (2017) 503–516.

558 [2] A. Aali, N. Pourmahmoud, V. Zare, Exergoeconomic analysis and multi-objective optimization
559 of a novel combined flash-binary cycle for Sabalan geothermal power plant in Iran, Energy
560 Conversion and Management 143 (2017) 377–390.

30
561 [3] D. Walraven, B. Laenen, W. D’haeseleer, Minimizing the levelized cost of electricity production
562 from low-temperature geothermal heat sources with ORCs: Water or air cooled?, Applied
563 Energy 142 (2015) 144–153.

564 [4] M. Usman, M. Imran, Y. Yang, D. H. Lee, B.-S. Park, Thermo-economic comparison of air-
565 cooled and cooling tower based Organic Rankine Cycle (ORC) with R245fa and R1233zde
566 as candidate working fluids for different geographical climate conditions, Energy 123 (2017)
567 353–366.

568 [5] D. Budisulistyo, C. S. Wong, S. Krumdieck, Lifetime design strategy for binary geothermal
569 plants considering degradation of geothermal resource productivity, Energy Conversion and
570 Management 132 (2017) 1–13.

571 [6] C. Yilmaz, Thermoeconomic cost analysis and comparison of methodologies for Dora II binary
572 geothermal power plant, Geothermics 75 (2018) 48–57.

573 [7] C. Clauser, M. Ewert, The renewables cost challenge: Levelized cost of geothermal electric
574 energy compared to other sources of primary energy Review and case study, Renewable and
575 Sustainable Energy Reviews 82 (2018) 3683–3693.

576 [8] T. T. Tran, A. D. Smith, Incorporating performance-based global sensitivity and uncertainty
577 analysis into LCOE calculations for emerging renewable energy technologies, Applied Energy
578 216 (2018) 157–171.

579 [9] K. Braimakis, S. Karellas, Integrated thermoeconomic optimization of standard and regener-
580 ative ORC for different heat source types and capacities, Energy 121 (2017) 570–598.

581 [10] C. Zhang, C. Liu, S. Wang, X. Xu, Q. Li, Thermo-economic comparison of subcritical organic
582 Rankine cycle based on different heat exchanger configurations, Energy 123 (2017) 728–741.

583 [11] H. Tian, L. Chang, Y. Gao, G. Shu, M. Zhao, N. Yan, Thermo-economic analysis of zeotropic
584 mixtures based on siloxanes for engine waste heat recovery using a dual-loop organic Rankine
585 cycle (DORC), Energy Conversion and Management 136 (2017) 11–26.

586 [12] H. Xi, M.-J. Li, Y.-L. He, Y.-W. Zhang, Economical evaluation and optimization of organic
587 Rankine cycle with mixture working fluids using R245fa as flame retardant, Applied Thermal
588 Engineering 113 (2017) 1056–1070.

31
589 [13] D. F. Dominković, M. Wahlroos, S. Syri, A. S. Pedersen, Influence of different technologies
590 on dynamic pricing in district heating systems: Comparative case studies, Energy 153 (2018)
591 136–148.

592 [14] M. Huculak, W. Jarczewski, M. Dej, Economic aspects of the use of deep geothermal heat in
593 district heating in Poland, Renewable and Sustainable Energy Reviews 49 (2015) 29–40.

594 [15] A. Keçeba, Energetic, exergetic, economic and environmental evaluations of geothermal district
595 heating systems: An application, Energy Conversion and Management 65 (2013) 546–556.

596 [16] M. Mehrpooya, M. Ashouri, A. Mohammadi, Thermoeconomic analysis and optimization of

597 a regenerative two-stage organic Rankine cycle coupled with liquefied natural gas and solar
598 energy, Energy 126 (2017) 899–914.

599 [17] E. Akrami, A. Chitsaz, H. Nami, S. Mahmoudi, Energetic and exergoeconomic assessment of a
600 multi-generation energy system based on indirect use of geothermal energy, Energy 124 (2017)
601 625–639.

602 [18] F. A. Boyaghchi, H. Safari, Parametric study and multi-criteria optimization of total exergetic
603 and cost rates improvement potentials of a new geothermal based quadruple energy system,
604 Energy Conversion and Management 137 (2017) 130–141.

605 [19] R. Leiva-Illanes, R. Escobar, J. M. Cardemil, D. C. Alarcón-Padilla, Comparison of the lev-

606 elized cost and thermoeconomic methodologies Cost allocation in a solar polygeneration plant
607 to produce power, desalted water, cooling and process heat, Energy Conversion and Manage-
608 ment 168 (2018) 215–229.

609 [20] S. Karellas, K. Braimakis, Energyexergy analysis and economic investigation of a cogeneration
610 and trigeneration ORCVCC hybrid system utilizing biomass fuel and solar power, Energy
611 Conversion and Management 107 (2016) 103–113.

612 [21] E. A. Pina, M. A. Lozano, L. M. Serra, Allocation of economic costs in trigeneration systems
613 at variable load conditions including renewable energy sources and thermal energy storage,
614 Energy 151 (2018) 633–646.

32
615 [22] A. S. Mundada, K. K. Shah, J. M. Pearce, Levelized cost of electricity for solar photovoltaic,
616 battery and cogen hybrid systems, Renewable and Sustainable Energy Reviews 57 (2016)
617 692–703.

618 [23] A. M. Pantaleo, S. M. Camporeale, A. Miliozzi, V. Russo, N. Shah, C. N. Markides, Novel

619 hybrid CSP-biomass CHP for flexible generation: Thermo-economic analysis and profitability
620 assessment, Applied Energy 204 (2017) 994–1006.

621 [24] E. Martelli, F. Capra, S. Consonni, Numerical optimization of Combined Heat and Power
622 Organic Rankine Cycles Part A: Design optimization, Energy 90 (2015) 310–328.

623 [25] M. Noussan, Allocation factors in Combined Heat and Power systems Comparison of different
624 methods in real applications, Energy Conversion and Management 173 (2018) 516–526.

625 [26] D. Walraven, B. Laenen, W. D’haeseleer, Optimum configuration of shell-and-tube heat ex-
626 changers for the use in low-temperature organic Rankine cycles, Energy Conversion and Man-
627 agement 83 (2014) 177–187.

628 [27] S. Van Erdeweghe, J. Van Bael, B. Laenen, W. D’haeseleer, Design and off-design opti-
629 mization procedure for low-temperature geothermal organic Rankine cycles (Working paper,
630 submitted for publication), 2018. URL: https://www.mech.kuleuven.be/en/tme/research/
631 energy_environment/Pdf/wp-en-2018-13-3.pdf.

632 [28] S. Bos, B. Laenen, Development of the first deep geothermal doublet in the Campine Basin of
633 Belgium, European Geologist 43 (2017) 16–20.

634 [29] CREG, Prijs van elektriciteit en aardgas in België, in de 3 regio’s en in de

635 buurlanden, 2018. URL: https://www.creg.be/sites/default/files/assets/Prices/
636 BelEnergyPriceCompNL.pdf.

637 [30] IRENA, Renewable energy in district heating and cooling - Case studies, 2017. URL: www.
638 irena.org/remap.

639 [31] P. Garg, M. S. Orosz, P. Kumar, Thermo-economic evaluation of ORCs for various working
640 fluids, Applied Thermal Engineering 109 (2016) 841–853.

33
641 [32] J. Calm, G. Hourahan, Physical, Safety and Environmental Data for Current
642 and Alternative Refrigerants, in: International Congress of Refrigeration, Prague,
643 Czech Republic, 2011. URL: http://www.hourahan.com/wp/wp-content/uploads/2010/08/
644 2011-Physical-Safety-and-Environmental-Data2.pdf.

645 [33] S. Van Erdeweghe, J. Van Bael, B. Laenen, W. D’haeseleer, Optimal configuration for a low-
646 temperature geothermal CHP plant based on thermoeconomic optimization (Working paper,
647 submitted for publication), 2018. URL: https://www.mech.kuleuven.be/en/tme/research/
648 energy_environment/Pdf/wp-en-2018-17.pdf.

649 [34] IEA, Technology Roadmap - Geothermal Heat and Power, 2011. URL:
650 https://www.iea.org/newsroomandevents/pressreleases/2011/june/
651 how-to-achieve-at-least-a-tenfold-increase-in-supply-of-geothermal-power-and-hea.
652 html.

653 [35] European Commission, EU Reference Scenario 2016, 2014. URL: https://ec.europa.eu/
654 energy/sites/ener/files/documents/ref2016_report_final-web.pdf.

655 [36] F. Heberle, D. Brüggemann, Thermoeconomic Analysis of Hybrid Power Plant Concepts for
656 Geothermal Combined Heat and Power Generation, Energies 7 (2014) 4482–4497.

34