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

Warmest vs WarmestTemperatureFlow SetpointManager

asked 2017-08-28 19:43:28 -0500

updated 2017-09-13 11:22:03 -0500

Molly Curtz gravatar image

I'm trying to understand the vast differences I get in energy usage between different SetpointManager options. This originated from a CBECC-Com model. From what I can tell from the I/O and Engineering References:

SetpointManager:Warmestresets the cooling SAT based on the cooling demand of the warmest zone, at the maximum zone supply air flow rate.

SetpointManager:WarmestTemperatureFlow "Temp First" resets the cooling SAT based on the cooling demand of the warmest zone, at the zone minimum supply air flow rate. If this SAT is less than the minimum, the SAT is set to the minimum, and the flow rate is increased until the cooling loads are met.

SetpointManager:WarmestTemperatureFlow "Flow First" resets the supply air flow rate based on the cooling demand of the warmest zone, at the zone maximum supply air temperature. If the flow is greater than the maximum, the flow is set to the maximum and the SAT is reduced until the cooling loads are met.

I tested all three in a project (from CBECC-Com) with a built-up VAV reheat system (water-cooled chiller, and boiler) and came up with this plot of HVAC energy usage. Everything in the models is hard-sized. I also plotted hourly results from the same node as the setpoint manager, but wasn't able to make sense of the results.

image description


  1. Why do these options yield such different results?

  2. In what scenarios would it make sense to implement the SetpointManager:WarmestTemperatureFlow Flow First or Temperature First options vs the SetpointManager:Warmest? I want to generally understand what the pros and cons are for each option for different system configurations.

edit retag flag offensive close merge delete


I added more details, mentioning that this is from a CBECC-Com model, so all inputs are hard-sized. There is no auto-sizing happening. It is also a built-up VAV system with a water-cooled chiller and boiler.

Anna Osborne Brannon gravatar imageAnna Osborne Brannon ( 2017-08-28 22:29:08 -0500 )edit

3 Answers

Sort by ยป oldest newest most voted

answered 2017-08-28 21:32:38 -0500

TempFirst will drive down the Supply Air Temperature as the first step in the control strategy. This will minimize fan energy, at the cost of greater reheat and chiller energy. Use this strategy for a Low temp VAV System. Low temperature refers the supply air temps 10 DBC or less to be true 'Low temp'.
The FlowFirst strategy will offload the chillers, at the expense of the AHU Fans.

The energy comparison you've shown may be skewed by the sizing details of the chilled water plant, cooling coils, AHU Fans, Supply air temperatures in the SetpointManager ranges.

For an effective Low Temp VAV system you need the chillers to be delivering cold water and the cooling coils to be sized for the corresponding temps. The chiller curves used could also confuse things - if you are extrapolating outside of the chilled water temperatures limits in the curves than the COPs could be erroneous. If you have hardsized any of these, than simply switching the SetpointManager may not lead to a fair comparison. Also, if you have autosized these, then you may not be operating the plant in the temperatures ranges to make the strategy efficient.

I would expect that the 2nd and 3rd scenarios should show roughly the same chiller+AHU energy, with chiller energy higher in the 2nd scenario. It would be interesting and perhaps helpful to see a plot (of one zone) Zone Air Temperature, Supply Air Temperature and Supply Air Volume Flow Rate for each scenario. While the Fan energy is lower in (2), as expected, the chiller energy is lower as well. That's the detail to focus on I reckon. The other big variable is the climate. The FlowFirst may make sense where the hours of economy cycle are high, and conversely, it could be a penalty in a hot and humid climate.

edit flag offensive delete link more


Very nice points.

Liam gravatar imageLiam ( 2017-08-28 21:40:59 -0500 )edit

Thanks for your points- I'll look into it a bit deeper and report back what I find.

Anna Osborne Brannon gravatar imageAnna Osborne Brannon ( 2017-08-28 22:11:59 -0500 )edit

What weather file was run? Did you verify that the max. And min. allowed reset limit on the SAT are the same in the three runs? What are the limits? What is the configuration of zones served? (e.g. Square floor plate with one core zone and 4 perimeter zones, 10 foot perimeter depth)

Molly Curtz gravatar imageMolly Curtz ( 2017-09-13 10:13:57 -0500 )edit

I second the point made in this answer that the chiller and chilled water loop control SPMs need to be considered. What you are doing on the airside may not be well coordinated with the waterside controls. For example, if your chilled water loop is constant volume flow with a scheduled SPM.

Molly Curtz gravatar imageMolly Curtz ( 2017-09-13 11:09:02 -0500 )edit

answered 2017-08-28 21:42:36 -0500

1) It will take more digging into the results to find the answer your looking for. These SPMs all set a temperature at the control node. Flow is not set by the Warmest FlowFirst manager, temperature is, which influences the request for flow by the zone terminal units. The documentation could have been a little more clear on that. What the Warmest and Warmest FlowFirst SPMs attempt to do is set the warmest zone terminal unit damper position to 1 while all other zones will have a damper position < 1. Damper position will affect how much reheat is needed.

Since these are cooling mode managers, the control node should be at the outlet of cooling coils only. Assuming the maximum zone air flow rate and min/max SAT are the same among these SPMs, then Warmest would have the highest SAT while Warmest TempFirst would have the lowest (with higher reheat requirement). I would think that Warmest FlowFirst would have similar results as Warmest since they both act to force zone air flow rate to the highest value. Your results indicate otherwise so I would be suspicious of these results.

I would look at the resulting zone temperature histograms from these runs. Also see how outdoor air loads (flow rate) change between these scenarios. It would also be nice to see a comparison of control node set point temperatures, supply air flow rates, damper positions, and reheat quantity.

Heat rejection did not change here so I wonder what heat transfer that variable represents.

2) I can only think of one reason to choose Warmest TempFirst and that would be humidity control. A histogram of zone relative humidity will show how these managers control latent loads. One reason I can think of to include TempFirst and FlowFirst as control types would be to switch between the two SPMs (Warmest and WarmestTemperatureFlow) without actually having to change objects (i.e., you can change between Warmest and Warmest TempFirst simply by changing the control type.

I would be interested in seeing more results to get a better idea of what is happening here.

edit flag offensive delete link more


Thanks for the comments and suggestions. I'll look more into it.

Anna Osborne Brannon gravatar imageAnna Osborne Brannon ( 2017-08-28 22:15:20 -0500 )edit

We should note that Warmest FlowFirst, and Warmest are not exactly the same. For Warmest, one critical connected zone will always be a maximum flow. Whereas, for Warmest Flow First, it is possible that all connected zones are satisfied at the maximum allowed setpoint temp at various flow rates below the maximum design flow. Only if the required flow for a zone at maximum setpoint temperature exceeds the maximum flow, will flow for that zone be set to maximum and the supply temp dropped (as for Warmest).

Molly Curtz gravatar imageMolly Curtz ( 2017-09-13 10:58:11 -0500 )edit

@rraustad, Thanks for clarifying that Flow rates are NOT set by the SPM:WarmestTemperaureFlow. I agree the documentation should be improved in the I/O reference to make that clear. Obviously flow rates are calculated by this SPM, but it's helpful to understand they are not SET.

Molly Curtz gravatar imageMolly Curtz ( 2017-09-13 11:14:09 -0500 )edit

answered 2017-08-28 21:39:50 -0500

Liam gravatar image

I think you are going to get different answers here from different people, but you seem to be asking a plant/system optimization question without providing details on the plant and system itself: what is your cooling source? Heating source? Pump sizing and configuration? What kind of fan do you have? What are your equipment curves?

I think you've asked a question with an open-ended answer because these will each be different based on your building dynamics - you may not have the same answer building to building or even for the same building with a different plant/system. But, that is why simulating them as you've done is important.

In an attempt to answer your question - you might use flow first where your flow curves are more efficient (i.e., fans) and you might use temperature first where your plant curves are more efficient (i.e., pumps).

edit flag offensive delete link more


Yes, it is open-ended, intentionally. I'm not trying to test this for a specific project, I'm just trying to understand the general differences for the control strategies. I find it useful to understand the concepts at a high level before trying to get too deep on implications for a specific system or project. But your point is taken that I should provide more information so that readers can make sense of and interpret the results I presented above.

Anna Osborne Brannon gravatar imageAnna Osborne Brannon ( 2017-08-28 22:11:14 -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



Asked: 2017-08-28 19:43:28 -0500

Seen: 91 times

Last updated: Sep 13