Geothermal combined heat and power plant design
A binary geothermal plant that adds electricity generation to a 50 MW district-heating supply, designed from first principles in MATLAB.
- Context
- TU/e coursework, Sustainable Energy Sources (4EM70) · 2026
- Role
- Group project
- Tools
- MATLAB, XSteam (IAPWS-IF97)
Problem
A heat-only plant uses 240 °C geothermal brine to supply a town's district-heating network with 50 MW at 110 °C. The brine is hot enough to generate electricity as well. In a binary combined heat and power (CHP) plant, the brine boils water in a separate steam cycle; the steam drives a turbine, and the turbine's condenser heats the district-heating water instead of rejecting that heat to the air.
The assignment was to design both plants and compare them: how much electricity the CHP plant adds, what equipment it needs, and how quickly the extra investment pays back.
My contribution
An individual coursework project following the course's design specification. I wrote the model in MATLAB, made and justified the design choices, sized the equipment and estimated the economics. The model checks its own results each time it runs.
Approach
Heat-only plant. A single counterflow shell-and-tube exchanger transfers 50 MW from the brine to the district-heating water, which returns at 26 °C and leaves at 110 °C.
Cycle. Water evaporates at 180 °C (10.0 bar) and condenses at 120 °C (2.0 bar). The condensing temperature sits 10 K above the 110 °C supply so the condenser can heat the network, and evaporating at 180 °C keeps the steam at the turbine exit about 90% dry without superheating.
Equipment. The economiser, evaporator and condenser are shell-and-tube exchangers sized down to tube count and length, with a 10 K minimum temperature difference. Single-phase sides use Dittus–Boelter heat-transfer coefficients; the boiling and condensing sides are treated as having negligible thermal resistance. Equipment costs come from the course's cost charts, based on the ICARUS data tabulated by Loh et al. (2002), in 1998 US$.
Checks. Each run evaluates 40 checks: energy balances on every exchanger and on the cycle, the Second Law, exchanger effectiveness against NTU theory, the validity ranges of the heat-transfer correlations, and design limits such as minimum temperature differences and turbine exit quality. If a critical check fails, the run ends with an error that marks the design as invalid. In the base design, all 40 pass.
Results
| Electricity | 7.2 MW | Cycle output: turbine 7.18 MW less feed pump 0.02 MW |
|---|---|---|
| District heating | 50 MW | Unchanged from the heat-only plant, supplied at 110 °C |
| Cycle efficiency | 12.5% | Against a Carnot limit of 13.2% between 180 °C and 120 °C |
| Steam flow | 25.2 kg/s | Brine flow 222 kg/s, reinjected at 183 °C |
| Exchanger areas | 666 m² | Evaporator 372, condenser 226 and economiser 67 m² |
Payback. The CHP plant's equipment (turbine and condenser, evaporator, economiser and feed pump) costs about US$1.21 million in purchased cost, which is US$1.19 million more than the heat-only plant's single exchanger. Selling the electricity at US$0.09/kWh, the estimated payback is about 1,840 operating hours. That is an estimate under the course's cost method: purchased carbon-steel equipment at 1998 prices, with installation, wells, drilling and the heat network excluded. By comparison, the heat-only exchanger pays back in about 16 hours.
Limitations
- The turbine and pump are ideal (isentropic), so the electrical output is an upper bound.
- The boiling and condensing sides of the exchangers are assumed to have negligible thermal resistance, so those exchanger areas are lower bounds.
- The payback uses purchased equipment cost only, from 1998 cost charts, with no pressure correction for the 50 bar brine side.
- Pumping power for the brine and the district-heating loop is not deducted from the electricity produced.
- The brine is modelled as pure water, and the reservoir temperature is assumed constant.