Question-and-Answer Resource for the Building Energy Modeling Community
Get s tarted with the Help page
Ask Your Question

Unmet Load Hours Linked to Cycling Run Time?

asked 2014-12-04 12:28:24 -0500

jdunn gravatar image

updated 2015-07-27 11:55:53 -0500

Hi All -

I'm investigating the inability of my EnergyPlus model to maintain setpoint during unoccupied hours.

The short story:

For my centralized VAV system (modeled through template objects), it appears that the AvailabilityManager:NightCycle control (cycleonany) is working properly for the time input into the "Cycling Run Time" field (currently set for 3,600 seconds - 1 hour). However, at the end of the cycle, the equipment drops to zero for one timestep, before cycling back on for an hour. Thus, the cooling coil energy will have values for 6 timesteps, then a zero, then values again for 6 timesteps, then zero, etc. This causes a drop in the VAV damper position and an increase supply air temperature (cooling) for said timestep, which allows the zone temp to float above the setback. I have no idea why it is doing this.

Image of a Sunday and Monday - notice the cycling of the AHU-2 cooling coil energy and VAV terminal damper temperature.

image description

The long story:

100+ zone, 2-story school. I used the HVACTemplateSystem:VAV, HVACTemplateZone:VAV, and constituent plant objects to model 6 air handling units for the various thermal zones of the building.

The zone I'm analyzing is an office zone, and I double checked all internal loads to make sure nothing was out of whack.

I've looked at the chilled water temperature leaving the plant and entering the air handling units, and the water temperatures look good. Its only the air temperatures leaving the coil that oscillate with the cycling behavior.

I've tried to change the "Cycling Run Time field: in the AvailabilityManager:NightCycle, but the mysterious zero values still happen at the end. It's as if it takes a timestep for the system to realized it needs to come back on. Increasing the cycling run time to like 9 hours keeps the temperature in range, but subcools the space as the equipiment will run for that defined period, no matter what the temperature is. I'm running the model at 6 timesteps, which is the only reason I saw the mysterious zero values in the first place.

Also, looking at the other 5 air handlers, they all exhibit the same cycling behavior - EXCEPT the gym zone. It is the only one that doesn't, but I'm fairly certain everything is modeled the same. Here's an image of all 8 air handling unit cooling coil energies, notice the cycling during the unoccupied times.

image description

Link to the expanded .idf file - set up to run for 1 week in June, 6 timesteps.

Link Hourly CSV File

edit retag flag offensive close merge delete

2 Answers

Sort by ยป oldest newest most voted

answered 2014-12-18 16:07:02 -0500

As you suggested in your question, the night cycle manager is always lagged by one time step, so the zone temperature must fall beyond the availability manager "Thermostat Tolerance" for one zone time step, and then the system will come on at the next time step and run for the specified Cycling Run Time. You thermostat tolerance is set to 0.2C which is also the same as the tolerance for the thermostat setpoint not met outputs. So, you can reduce the number of unoccupied hours not met, but making the Thermostat Tolerance smaller in the availability manager, say 0.1C. But there may still be times where the space is heating up fast enough to be 0.2 or more beyond setpoint in a single timestep.

edit flag offensive delete link more

answered 2014-12-05 10:17:30 -0500

Archmage gravatar image

Check the mass flow at the terminal during the off cycle -- sometime these fast ramp rates are caused because the system is still on and delivering unconditioned outside air into the zone (when you think it should be off).

When I look at these plots I see zones with too much load for the system to really be turned off. And typically the setup thermostat temperature for cooling during unoccupied times would be higher than 25C . But otherwise this is pretty much what happens with that manager when the loads are high. If the zones cannot handle cycling then leave the air handler running as normal.

edit flag offensive delete link more



Thanks for the reply!

Looking back at the mass flow rate and the outdoor air flow rate through the terminal, they both drop to 0 when the unit is cycled off, or during periods of no load during unoccupied times.

It does seem like the loads are too high, but upon investigation all load are within reason, and the envelope was specified correctly. The zone analyzed in the charts above is a pretty small private office, but this is also happening in the classroom spaces throughout the project. Something must be specified incorrectly and I'm just not seeing it....

jdunn gravatar image jdunn  ( 2014-12-11 09:04:48 -0500 )edit

I'd guess it is solar gains.

Archmage gravatar image Archmage  ( 2014-12-11 09:44:07 -0500 )edit

That's a good guess, but the window is facing north....

jdunn gravatar image jdunn  ( 2014-12-15 12:56:21 -0500 )edit

In general, I would suggest reporting the following set of variables to investigate the relative magnitudes of loads on the zone air.

Zone Air Heat Balance Internal Convective Heat Gain Rate [W] Zone Air Heat Balance Surface Convection Rate [W] Zone Air Heat Balance Interzone Air Transfer Rate [W] Zone Air Heat Balance Outdoor Air Transfer Rate [W] Zone Air Heat Balance System Air Transfer Rate [W] Zone Air Heat Balance System Convective Heat Gain Rate [W] Zone Air Heat Balance Air Energy Storage Rate [W]

Archmage gravatar image Archmage  ( 2014-12-15 13:31:50 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer


Question Tools

1 follower


Asked: 2014-12-04 12:28:24 -0500

Seen: 283 times

Last updated: Dec 18 '14