Build a Heat Pump Stepwise#

We provide the full script presented in this tutorial here: stepwise.py

Figure: Topology of the heat pump system#

Task#

This tutorial introduces you in how to model a heat pump in TESPy. You can see the plant’s topology in the figure.

The main purpose of the heat pump is to deliver heat e.g. for the consumers of a heating system. Thus, the heat pump’s parameters will be set in a way, which supports this target. Generally, if systems are getting more complex, it is highly recommended setting up your plant in incremental steps. This tutorial divides the plant in three sections: The consumer part, the valve and the evaporator and the compressor as last element. Each new section will be appended to the existing ones.

The system will be built up in a way, that independent of what working fluid we use, we will be able to generate stable starting values. After achieving that, the final parameters are specified instead of the initial values.

Set up a Network#

First, we have to create an instance of the tespy.networks.network.Network class. The network is the main container of the model and will be required in all following sections. In this example we will work with water (H₂O) and ammonia (NH₃). Water is used for the cold side of the heat exchanger, for the consumer and for the hot side of the environmental temperature. Ammonia is used as refrigerant within the heat pump circuit. If you don’t specify the unit system, the variables are set to SI-Units. We also keep the working fluid a variable to make reusing the script with a different working fluid easy.

from tespy.networks import Network
working_fluid = "NH3"

nw = Network(
    T_unit="C", p_unit="bar", h_unit="kJ / kg", m_unit="kg / s"
)

We will use °C, bar and kJ/kg as units for temperature and enthalpy.

Modeling the heat pump: Consumer system#

Components#

We will start with the consumer as the plant will be designed to deliver a specific heat flow. From the figure above you can determine the components of the consumer system: condenser, pump and the consumer (tespy.components.heat_exchangers.simple.SimpleHeatExchanger). Additionally, we need a source and a sink for the consumer and the heat pump circuit respectively. We will import all necessary components already in the first step, so the imports will not need further adjustment.

Tip

We label the sink for the refrigerant "valve", as for our next calculation the valve will be attached there instead of the sink. In this way, the fluid properties can be initialized at the interface-connection, too.

from tespy.components import Condenser
from tespy.components import CycleCloser
from tespy.components import SimpleHeatExchanger
from tespy.components import Pump
from tespy.components import Sink
from tespy.components import Source

# sources & sinks
c_in = Source("refrigerant in")
cons_closer = CycleCloser("consumer cycle closer")
va = Sink("valve")

# consumer system
cd = Condenser("condenser")
rp = Pump("recirculation pump")
cons = SimpleHeatExchanger("consumer")

Connections#

In the next steps we will connect the components in order to form the network. Every connection requires the source, the source id, the target and the target id as arguments: the source is the component from which the connection originates, the source id is the outlet id of that component. This applies analogously to the target. To find all inlet and outlet ids of a component look up the class documentation of the respective component. An overview of the components available and the class documentations is provided in the TESPy modules overview. The tespy.connections.connection.Ref class is used specify fluid property values by referencing fluid properties of different connections. It is used in a later step.

from tespy.connections import Connection

c0 = Connection(c_in, "out1", cd, "in1", label="0")
c1 = Connection(cd, "out1", va, "in1", label="1")

c20 = Connection(cons_closer, "out1", rp, "in1", label="20")
c21 = Connection(rp, "out1", cd, "in2", label="21")
c22 = Connection(cd, "out2", cons, "in1", label="22")
c23 = Connection(cons, "out1", cons_closer, "in1", label="23")

nw.add_conns(c0, c1, c20, c21, c22, c23)

Parametrization#

For the condenser we set pressure ratios on hot and cold side. The consumer will have pressure losses, too. Further we set the isentropic efficiency for the pump. The most important parameter is the consumer’s heat demand since it decides the overall mass flow in the systems.

Tip

In this tutorial we will first build the system with parameters that ensure stable starting values for a simulation, which in the end will be switched to reasonable values for the individual parts of the system. For example, instead of the evaporation pressure we will use the terminal temperature difference at the condenser instead.

cd.set_attr(pr1=0.99, pr2=0.99)
rp.set_attr(eta_s=0.75)
cons.set_attr(pr=0.99)

In order to calculate this network further parametrization is necessary, as e.g. the fluids are not determined yet: At the hot inlet of the condenser we define the temperature, pressure and the fluid informaton. A good guess for pressure can be obtained from CoolProp’s PropsSI function. We know that the condensation temperature must be higher than the consumer’s feed flow temperature. Therefore, we can set the pressure to a slightly higher value of that temperature’s corresponding condensation pressure.

The same needs to be done for the consumer cycle. We suggest setting the parameters at the pump’s inlet. On top, we assume that the consumer requires a constant inlet temperature. The CycleCloser automatically makes sure, that the fluid’s state at the consumer’s outlet is the same as at the pump’s inlet.

from CoolProp.CoolProp import PropsSI as PSI
p_cond = PSI("P", "Q", 1, "T", 273.15 + 95, working_fluid) / 1e5
c0.set_attr(T=170, p=p_cond, fluid={working_fluid: 1})
c20.set_attr(T=60, p=2, fluid={"water": 1})
c22.set_attr(T=90)

# key design paramter
cons.set_attr(Q=-230e3)

Solve#

After creating the system, we want to solve our network. Until we have not set up the full system we will run design case calculations.

Note

In TESPy there are two different types of calculations: design point and offdesign calculation.

Generally, the design calculation is used for designing your system in the way you want it to look like. This means, that you might want to specify a design point isentropic efficiency, pressure loss or terminal temperature difference. After you have designed your system, you are able to make offdesign calculations with TESPy. The offdesign calculation is used to predict the system’s behavior at different points of operation. For this case, this might be different ambient temperature, different feed flow temperature, or partial load. Add the end of this tutorial, you will learn how to run the offdesign calculation.

Valve and evaporator system#

Next we will add the valve and the evaporator system to our existing network. The figure below indicates the sections we will append in this step. This part contains of a valve followed by an evaporator with a drum (separating saturated liquid from saturated gas) and a superheater.

Components#

First, we need to import the new components, which are tespy.components.nodes.drum.Drum, tespy.components.heat_exchangers.base.HeatExchanger and tespy.components.piping.valve.Valve. We will add these components to the script.

from tespy.components import Valve, Drum, HeatExchanger

# ambient heat source
amb_in = Source("source ambient")
amb_out = Sink("sink ambient")

# evaporator system
va = Valve("valve")
dr = Drum("drum")
ev = HeatExchanger("evaporator")
su = HeatExchanger("superheater")

# virtual source
cp1 = Sink("compressor 1")

Connections#

Since the old connection 1 lead to a sink, we have to replace this connection in the network. We can do that by using the method del_conns passing c1. After that, we can create the new connections and add them to the network as we did before.

The valve connects to the drum at the inlet 'in1'. The drum’s outlet 'out1' is saturated liquid and connects to the evaporator’s cold side inlet 'in2'. The inlet reconnects to the drum’s inlet 'in2'. The superheater’s cold side connects to the drum’s outlet 'out2'. On the ambient side we simply connect the source to the superheater to the evaporator and finally to the ambient sink. This will add the following connections to the model:

nw.del_conns(c1)

# evaporator system
c1 = Connection(cd, "out1", va, "in1", label="1")
c2 = Connection(va, "out1", dr, "in1", label="2")
c3 = Connection(dr, "out1", ev, "in2", label="3")
c4 = Connection(ev, "out2", dr, "in2", label="4")
c5 = Connection(dr, "out2", su, "in2", label="5")
c6 = Connection(su, "out2", cp1, "in1", label="6")

nw.add_conns(c1, c2, c3, c4, c5, c6)

c17 = Connection(amb_in, "out1", su, "in1", label="17")
c18 = Connection(su, "out1", ev, "in1", label="18")
c19 = Connection(ev, "out1", amb_out, "in1", label="19")

nw.add_conns(c17, c18, c19)

Attention

The drum is special component, it has an inbuilt CycleCloser, therefore, although we are technically forming a cycle at the drum’s outlet 1 to its inlet 2, we do not need to include a CycleCloser here.

Parametrization#

Previous parametrization stays untouched. Regarding the evaporator, we specify pressure ratios on hot side as well as the evaporation pressure, for which we can obtain a good initial guess based on the ambient temperature level using CoolProp. From this specification the pinch point layout will be a result, similar as in waste heat steam generators. The pressure ratio of the cold side MUST NOT be specified in this setup as the drum assumes pressure equality for all inlets and outlets.

The superheater will also use the pressure ratios on hot and cold side. Further we set a value for the enthalpy at the working fluid side outlet. This determines the degree of overheating and is again based on a good guess.

ev.set_attr(pr1=0.99)
su.set_attr(pr1=0.99, pr2=0.99)

Next step is the connection parametrization: The pressure in the drum and the enthalpy of the wet steam reentering the drum need to be determined. For the enthalpy we can specify the vapor mass fraction x determining the degree of evaporation. On the hot side inlet of the superheater we define the temperature, pressure and the fluid. At last, we have to fully determine the state of the incoming fluid at the superheater’s hot side.

# evaporator system cold side
c4.set_attr(x=0.9, T=5)

h_sat = PSI("H", "Q", 1, "T", 273.15 + 15, working_fluid) / 1e3
c6.set_attr(h=h_sat)

# evaporator system hot side
c17.set_attr(T=15, fluid={"water": 1})
c19.set_attr(T=9, p=1.013)

Solve#

We can again run a simulation after adding these parts. This step is not required, but in larger, more complex networks, it is recommended to achieve better convergence.

nw.solve("design")
nw.print_results()

Compressor system#

To complete the heat pump, we will add the compressor system to our existing network. This requires to change the connections 0, 6 and 17. The connection 6 has to be changed to include the compressor. After the last compressor stage, connection 0 has to redefined, since we need to include the CycleCloser of the working fluid’s cycle. The connection 17 has to be connected to the heat exchanger for intermittent cooling as well as the bypass.

Components#

This part contains two compressors with intermittent cooling between them. The cold side of the heat exchanger will be connected to a pump upstream and to the superheater downstream. The bypass is used to give the system flexibility in the temperature levels between the heat exchangers. We will also replace the source for the refrigerant of c0 at the condenser with another CycleCloser to make sure the fluid properties after the second compressor are identical to the fluid properties at the condenser’s inlet.

Tip

The intermittent cooling extracts heat from the cycle. As this heat is however used to increase the evaporation pressure of the working fluid due to the higher temperature level of the heat source, the reduction is very limited. We use a two stage compressor, because in a single stage compression, the outlet temperature of the refrigerant might violate technical boundary conditions of the real-world component.

from tespy.components import Compressor, Splitter, Merge

cp1 = Compressor("compressor 1")
cp2 = Compressor("compressor 2")

ic = HeatExchanger("intermittent cooling")
hsp = Pump("heat source pump")

sp = Splitter("splitter")
me = Merge("merge")
cv = Valve("control valve")

hs = Source("ambient intake")
cc = CycleCloser("heat pump cycle closer")

Connections#

We remove connections 0, 6 and 13 from the network, define the new connections and add them again.

nw.del_conns(c0, c6, c17)

c6 = Connection(su, "out2", cp1, "in1", label="6")
c7 = Connection(cp1, "out1", ic, "in1", label="7")
c8 = Connection(ic, "out1", cp2, "in1", label="8")
c9 = Connection(cp2, "out1", cc, "in1", label="9")
c0 = Connection(cc, "out1", cd, "in1", label="0")

c11 = Connection(hs, "out1", hsp, "in1", label="11")
c12 = Connection(hsp, "out1", sp, "in1", label="12")
c13 = Connection(sp, "out1", ic, "in2", label="13")
c14 = Connection(ic, "out2", me, "in1", label="14")
c15 = Connection(sp, "out2", cv, "in1", label="15")
c16 = Connection(cv, "out1", me, "in2", label="16")
c17 = Connection(me, "out1", su, "in1", label="17")

nw.add_conns(c6, c7, c8, c9, c0, c11, c12, c13, c14, c15, c16, c17)

Parametrization#

For the first compressor we set the pressure ratio to the square root of the full pressure ration between condensation and evaporation. In the first step, we do not set the isentropic efficiency, because the respective equations are quite sensitive to good starting value. We will set these values after the full system has been calculated. The pump’s isentropic efficiency value is not as critical, therefore we set this value. The intermittent cooling causes pressure losses on both sides.

pr = (c1.p.val / c5.p.val) ** 0.5
cp1.set_attr(pr=pr)
ic.set_attr(pr1=0.99, pr2=0.98)
hsp.set_attr(eta_s=0.75)

Regarding the connections we set enthalpy values for all working fluid side connections. After the superheater and intermittent cooling the value will be near saturation (enthalpy value of connection c5), after the compressors it will be higher.

For the ambient side, we set temperature, pressure and fluid at connection 11. On top of that, we can specify the temperature of the ambient water after leaving the intermittent cooler.

With re-adding of connection 0 we have to set the fluid and the pressure again, but not the temperature value, because this value will be a result of the condensation pressure and the given enthalpy at the compressor’s outlet.

c0.set_attr(p=p_cond, fluid={working_fluid: 1})

c6.set_attr(h=c5.h.val + 10)
c8.set_attr(h=c5.h.val + 10)

c7.set_attr(h=c5.h.val * 1.2)
c9.set_attr(h=c5.h.val * 1.2)

c11.set_attr(p=1.013, T=15, fluid={"water": 1})
c14.set_attr(T=30)

Solve and Set Final System Parameters#

Now we solve again. After that, we can exchange our good guesses with actual useful parameters:

The condensation and evaporation pressure levels will be replaced by terminal temperature values of the condenser and the evaporator respectively. The lower terminal temperature value of the evaporator ttd_l defines the pinch point. The upper terminal temperature value ttd_u of the condenser defines the condensation pressure.

The degree of superheating in the superheater will be determined by the upper terminal temperature instead of the enthalpy value at connection 6. The outlet enthalpies after both compressors are replaced by the isentropic efficiency values. Finally, the enthalpy after the intermittent cooling is replaced by the temperature difference to the boiling point. With this we can ensure, the working fluid does not start to condensate at the intermittent cooler.

nw.solve("design")

c0.set_attr(p=None)
cd.set_attr(ttd_u=5)

c4.set_attr(T=None)
ev.set_attr(ttd_l=5)

c6.set_attr(h=None)
su.set_attr(ttd_u=5)

c7.set_attr(h=None)
cp1.set_attr(eta_s=0.8)

c9.set_attr(h=None)
cp2.set_attr(eta_s=0.8)

c8.set_attr(h=None, Td_bp=4)
nw.solve("design")
nw.save("system_design")

Calculate Part Load Performance#

After setting up the full system, we want to predict part load operation at different values for the consumer’s heat demand. Some values utilized in the previous setup will change, if a component is not operated at its design point. This is individual to every system, so the designer has to answer the question: Which parameters are design point parameters and how does the component perform at a different operation point.

Tip

To make the necessary changes to the model, we can specify a design and an offdesign attribute, both lists containing component or connection parameter names. All parameters specified in the design attribute of a component or connection, will be unset in an offdesign calculation, all parameters specified in the offdesign attribute of a component or connection will be set for the offdesign calculation. The value for these parameters is the value derived from the design-calculation.

The changes we want to apply can be summarized as follows:

All heat exchangers should be calculated based on their heat transfer coefficient with a characteristic for correction of that value depending on the change of mass flow (kA_char). Therefore, terminal temperature value specifications need to be added to the design parameters. Also, the temperature at connection 14 cannot be specified anymore, since it will be a result of the intermittent cooler’s characteristics.
Pumps and compressors will have a characteristic function for their isentropic efficiency instead of a constant value (eta_s_char).
Pressure drops in components will be a result of the changing mass flow through that component given the diameter in the design. The pressure ratio will therefore be replaced by zeta for all heat exchangers. The zeta value is a geometry independent value.

On top of that, for the evaporator the characteristic function of the heat transfer coefficient should follow different data than the default characteristic. The name of that line is ‘EVAPORATING FLUID’ for the cold side. The default line ‘DEFAULT’ will be kept for the hot side. These lines are available in the tespy.data module.

Attention

If you run the offdesign simulation without any changes in the specification values, the results must be identical as in the respective design case! If they are not, it is likely, something went wrong.

cp1.set_attr(design=["eta_s"], offdesign=["eta_s_char"])
cp2.set_attr(design=["eta_s"], offdesign=["eta_s_char"])
rp.set_attr(design=["eta_s"], offdesign=["eta_s_char"])
hsp.set_attr(design=["eta_s"], offdesign=["eta_s_char"])

cons.set_attr(design=["pr"], offdesign=["zeta"])

cd.set_attr(
    design=["pr2", "ttd_u"], offdesign=["zeta2", "kA_char"]
)

from tespy.tools.characteristics import CharLine
from tespy.tools.characteristics import load_default_char as ldc

kA_char1 = ldc("heat exchanger", "kA_char1", "DEFAULT", CharLine)
kA_char2 = ldc("heat exchanger", "kA_char2", "EVAPORATING FLUID", CharLine)
ev.set_attr(
    kA_char1=kA_char1, kA_char2=kA_char2,
    design=["pr1", "ttd_l"], offdesign=["zeta1", "kA_char"]
)

su.set_attr(
    design=["pr1", "pr2", "ttd_u"], offdesign=["zeta1", "zeta2", "kA_char"]
)

ic.set_attr(
    design=["pr1", "pr2"], offdesign=["zeta1", "zeta2", "kA_char"]
)
c14.set_attr(design=["T"])
nw.solve("offdesign", design_path="system_design")
nw.print_results()