There are 2 issues here.
The first being the lack of water flow data in heating mode.
- If you have data for cooling mode, you could use the same values for heating mode.
- I would review the manufacturers performance data again to be sure heating water flow rates are not mentioned somewhere in the documentation.
The second issue is the use of only a single value for any variable in these tables, in your case the water flow rates.
- The coefficients for these variables will not regress and will probably wind up as useless if not 0.
- The system should use these flow rates as constants during the simulation (no variation at all) since the regression does not include any variation in these values.