solar collector with parallel arrangement compute wrong pressure drop

[mwetter]

@jelgerjansen reports:

Problem description

If multiple panels are placed in parallel, something is wrong in the flow rate-pressure equations. If we simulate for example Buildings.Fluid.SolarCollectors.Examples.FlatPlate, one time with 1 panel, a 2nd time having 2 panels in parallel. One would expect that in the given setting (fixed pressure difference over the collectors), the flow rate would be twice as high in the situation of 2 panels compared to 1 panel. The simulation results however give the same mass flow rate.

Problem cause

I think the nominal mass flow rate that is used in Buildings.Fluid.SolarCollectors.BaseClasses.PartialSolarCollector to calculate the flow resistance is not correctly defined if multiple panels are in parallel.

Right now the code reads as follows:

model PartialSolarCollector "Partial model for solar collectors"
 extends Buildings.Fluid.Interfaces.LumpedVolumeDeclarations;
  extends Buildings.Fluid.Interfaces.TwoPortFlowResistanceParameters(final dp_nominal = dp_nominal_final);
  extends Buildings.Fluid.Interfaces.PartialTwoPortInterface(
    final m_flow_nominal=perPar.mperA_flow_nominal*perPar.A);

where

  final parameter Modelica.Units.SI.PressureDifference dp_nominal_final(
      displayUnit="Pa") = if sysConfig == Buildings.Fluid.SolarCollectors.Types.SystemConfiguration.Series
     then nPanels_internal*perPar.dp_nominal else perPar.dp_nominal
    "Nominal pressure loss across the system of collectors";

For serial configurations, the increased pressure drop is correctly accounted for. For parallel flows however, the nominal mass flow rate is not corrected: the mass flow rate per panel equals m_flow_nominal, but the total nominal mass flow rate (that results in a pressure drop dp_nominal_final) is equal to m_flow_nominal*nPanels_internal.

Problem solution

I think the following code is more correct:

model PartialSolarCollector "Partial model for solar collectors"
  extends Buildings.Fluid.Interfaces.LumpedVolumeDeclarations;
  extends Buildings.Fluid.Interfaces.TwoPortFlowResistanceParameters(final dp_nominal = dp_nominal_final);
  extends Buildings.Fluid.Interfaces.PartialTwoPortInterface(
    final m_flow_nominal=m_flow_nominal_final);

where

  final parameter Modelica.Units.SI.PressureDifference dp_nominal_final(
      displayUnit="Pa") = if sysConfig == Buildings.Fluid.SolarCollectors.Types.SystemConfiguration.Series
     then nPanels_internal*perPar.dp_nominal else perPar.dp_nominal
    "Nominal pressure loss across the system of collectors";

  final parameter Modelica.Units.SI.MassFlowRate m_flow_nominal_final(
      displayUnit="kg/s") = if sysConfig == Buildings.Fluid.SolarCollectors.Types.SystemConfiguration.Parallel
     then nPanels_internal*perPar.mperA_flow_nominal*perPar.A else perPar.mperA_flow_nominal*perPar.A
    "Nominal mass flow rate through the system of collectors";

These changes result in the following simulation results of Buildings.Fluid.SolarCollectors.Examples.FlatPlate for 1 and 2 panels (in parallel), respectively:

The mass flow rate is now double as high, which is to be expected for a constant pressure raise over two parallel instead of one single collector.

[mwetter]

@jelgerjansen: This is indeed a bug. I will correct it on

• maint_9.1.x (incl. revision notes)
• maint_10.0.x (incl. revision notes)
• master. On the master branch, the parameter computeFlowResistance should be removed as it is not used.

branch.

This will change the results of Examples.Tubular which uses sysConfig = Buildings.Fluid.SolarCollectors.Types.SystemConfiguration.Parallel