Comparison of Radiant Floor Heating vs Hot Water Heating Coil

asked 2016-03-26 20:05:20 -0500

RChidwick

509 ●2 ●9

updated 2016-03-27 06:41:03 -0500

__AmirRoth__

4436 ●5 ●17 http://bleedinggreenna...

I created a simulation of an existing building in OpenStudio 1.10.0. The heating in the simulation is provided by a hot water heating coil (which is supplied by a natural gas boiler). I have hourly gas data from the actual building which I am using to verify that my heating system is working correctly. The real building uses 'rail heating' - exposed pipes that sit about 6 inches off the floor. Hot water circulates through the pipes to provide radiant heating. I have tried to model this using the ZoneHVAC:LowTemperatureVariableFlow object in OpenStudio but the gas usage is way off, about 3 to 5 times greater than the usage from the actual building.

I thought I would try using the Coil:Heating:Water object in place of the ZoneHVAC:LowTemperatureVariableFlow object. I made an air loop with just a fan, the coil, an uncontrolled diffuser and the thermal zone that the coil is supplying. I used the SetpointManager:SingleZone:Reheat setpoint manager for the air loop. The gas usage profile shape was nearly an exact match to the profile from the actual building, but the values were several times less than the actual values (peak hourly usage from the simulation was 3.8 GJ vs 16 GJ for the actual data).

I'm wondering if it makes sense that the gas usage from the simulation using a heating coil is much less than the gas usage from the building which is essentially using a radiant floor?

Thank you.

edit retag flag offensive close merge delete

add a comment

answered 2016-03-26 20:26:57 -0500

Julien Marrec

29017 ●85 ●74 http://www.effibem.com/

First, if I understand correctly the heating system you are describing, this is merely a baseboard. A baseboard is typical a pipe with fins to enhance the heat transfer by having a greater surface area, but not necessarily: if you have enough piping itself there's no need for fins.

Second, I'd suggest you calibrate first on gas usage (energy) and not peak (demand). Make sure that you can get close enough, that will help you tune the assumptions you had to take (occupancy, actual R-values, average thermostat setpoints, infiltration levels etc).

Then you can consider trying to further refine to approximate the peak. But note that it will require a much deeper granularity in your modeling. You really have to understand how the heating system as a whole is controlled and translate that properly. For example it's likely that your hourly peak value could be on a cold morning if the building has temperature setbacks and "boost" mode to bring back the building at the right temperature.

Comments