Hydraulic simulation of circulation pump in combination with other circulation pumps or ext_grids

[dlohmeier]

I built a test for running a simulation of a district heating grid model with a circulation pump and an external grid:

net = pandapipes.create_empty_network("net", add_stdtypes=False)

j1, j2, j3, j4, j5, j6, j7 = pandapipes.create_junctions(net, 7, pn_bar=5, tfluid_k=283.15)

pandapipes.create_pipes_from_parameters(net, [j1, j2, j4, j5], [j2, j3, j5, j6], k_mm=1.,
                                        length_km=[0.4338, 0.2637, 0.2894, 0.427],
                                        diameter_m=0.1022)

pandapipes.create_heat_exchanger(net, j3, j4, diameter_m=0.1022, qext_w=200000)
pandapipes.create_heat_exchanger(net, j2, j5, diameter_m=0.1022, qext_w=200000)

pandapipes.create_fluid_from_lib(net, "water", overwrite=True)

pandapipes.create_pipes_from_parameters(
    net, [0, j7], [j7, 1], k_mm=1., length_km=0.213, diameter_m=0.1022
)
pandapipes.create_ext_grid(net, j4, p_bar=4, type="p")
pandapipes.create_source(net, j7, mdot_kg_per_s=1)

net_cpm = copy.deepcopy(net)
net_eg = copy.deepcopy(net)

pandapipes.create_circ_pump_const_mass_flow(net_cpm, 6, 0, p_flow_bar=5, mdot_flow_kg_per_s=5,
                                            t_flow_k=300, type='pt')

pandapipes.create_ext_grid(net_eg, 0, p_bar=5, t_k=300, type="pt")
pandapipes.create_sink(net_eg, 6, mdot_kg_per_s=5)

for nt in [net_cpm, net_eg]:
    pandapipes.pipeflow(nt, stop_condition="tol", iter=10, friction_model="nikuradse",
                        mode="all", transient=False, nonlinear_method="automatic",
                        tol_p=1e-4, tol_v=1e-4, use_numba=use_numba)

assert np.allclose(net_cpm.res_junction.values, net_eg.res_junction.values)
assert np.allclose(net_cpm.res_pipe.values, net_eg.res_pipe.values)
assert np.allclose(net_cpm.res_heat_exchanger.values, net_eg.res_heat_exchanger.values)
assert np.allclose(net_cpm.res_flow_control.values, net_eg.res_flow_control.values)
assert np.allclose(net_cpm.res_ext_grid.values[:1], net_eg.res_ext_grid.values[:1])

# this last test currently fails!
cp_junc = net_cpm.circ_pump_mass.flow_junction.values[0]
cp_mdot = net_cpm.circ_pump_mass.mdot_flow_kg_per_s.values[0]
assert np.isclose(net_cpm.res_pipe.loc[
                      (net_cpm.pipe.from_junction == cp_junc)
                      | (net_cpm.pipe.to_junction == cp_junc)].mdot_from_kg_per_s.sum(),
                  cp_mdot)

The last assertion fails as we cannot ensure that the mass flow through fed in by the circulation pump is the same as the one that enters it. In order to do so, we need to change the way we model the circulation pump. We need the following equations:

1. At the outlet node: p = p_flow_bar
2. At the outlet node: sum(mdot)_in = sum(mdot)_out
3. At the inlet node: sum(mdot)_in = sum(mdot)_out
4. Through the pump: mdot = mdot_flow_kg_per_s (or the same as v)

These equations are valid for the circulation pump with constant mass flow. For the other one, I am not sure yet. Currently, this seems to me like 4 equations for 3 elements (2 nodes, 1 branch). But maybe there still is a way to get this done within the current structure. One possibility would be to add an internal auxiliary node, so that the circulation pump could internally be seen as a flow controller and a pressure controller in series. Otherwise, it would be important to be able to make such changes inside the model.