Small correction in heat consumer for mode MF_TR

[dlohmeier]

the return temperature of the heat consumer is not fixed for the solver anymore, but calculated based on an exact estimate of qext from inlet and controlled return temperature

[JonasPfeiffer123]

Dear @dlohmeier,

I'm currently exploring the new return temperature control feature for heat_consumer, introduced in v0.11. Previously, I utilized a custom controller to adjust the controlled_mass_flow in heat_consumer based on a given output temperature. Now, I’m transitioning to the new approach, but my calculations fail under certain configurations.

I examined the provided functionality and corresponding tests in src/pandapipes/test/api/test_components/test_heat_consumer.py. While experimenting, I observed the following:

1. The original example works as expected, where one heat consumer is assigned treturn_kand the other controlled_mdot_kg_per_s.
2. Both heat consumers specifying controlled_mdot_kg_per_s(as in previous versions) also works fine.
3. However, when both heat consumers are assigned only treturn_k, the bidirectional calculation fails. Using the sequential mode also does not work. For clarity, qext_wis defined for both heat consumers.

Below is a minimal example demonstrating the issue:

def test_heat_consumer_result_extraction
    net = pp.create_empty_network("net", add_stdtypes=False, fluid="water")

    juncs = pp.create_junctions(
        net,
        nr_junctions=6,
        pn_bar=5,
        tfluid_k=286,
        system=["flow"] * 3 + ["return"] * 3,
        geodata=[
            (0, 0),    # Junction 0 (Startpunkt)
            (10, 0),   # Junction 1 (Vorlauf)
            (20, 0),   # Junction 2 (Vorlauf)
            (20, -10), # Junction 3 (Rücklauf)
            (10, -10), # Junction 4 (Rücklauf)
            (0, -10),  # Junction 5 (Endpunkt Rücklauf)
        ],
        name=[
            "Junction 0", # Junction 0
            "Junction 1", # Junction 1
            "Junction 2", # Junction 2
            "Junction 3", # Junction 3
            "Junction 4", # Junction 4
            "Junction 5" # Junction 5
        ]
    )
    pp.create_pipes_from_parameters(net, juncs[[0, 1, 3, 4]], juncs[[1, 2, 4, 5]], k_mm=0.1, length_km=1,
                                            diameter_m=0.1022, system=["flow"] * 2 + ["return"] * 2, alpha_w_per_m2k=0.4,
                                            text_k=273.15, name=[
            "Flow Pipe 1",  # Pipe von Source zu Flow Node 1
            "Flow Pipe 2",  # Pipe von Flow Node 1 zu Flow Node 2
            "Return Pipe 1", # Pipe von Return Node 1 zu Return Node 2
            "Return Pipe 2"  # Pipe von Return Node 2 zu Sink
        ])
    pp.create_circ_pump_const_pressure(net, juncs[-1], juncs[0], 5, 2, 85+273.15, type='auto', name="Pump 1")
    pp.create_heat_consumer(net, juncs[1], juncs[4], treturn_k=60+273.15, qext_w=7500, name="Heat Consumer 1")
    pp.create_heat_consumer(net, juncs[2], juncs[3], treturn_k=75+273.15, qext_w=7500, name="Heat Consumer 2")

    #hydraulics only to check for lookup heat transfer error
    pp.pipeflow(net, iter=3)

    pp.pipeflow(net, mode="bidirectional", iter=50) # played around with iter

    return net

Observations:

When both HeatConsumers are set to treturn_kwith a defined qext_w, the bidirectional calculation fails. From my understanding of the new implementation, it should be possible to define output temperatures for all HeatConsumers without needing to set controlled_mdot_kg_per_s. Is my assumption correct? Or is at least one heat consumer with a set mass flow (controlled_mdot_kg_per_s) required?

Thank you in advance for your clarification and any advice you can provide!