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

# Meeting Humidity Setpoint with Evaporative Media

I am setting up an HVAC system for a computer room. The person density in the computer room is 0 and the computers are always on generating much heat. There is no outdoor airflow due to fear of particles, and the cooling coil is always on, day and night, which condenses much of the water out of the air. Low humidity is bad for computers since static charge can build up and damage components, so water must be added to the air in the loop.

The best objects I've found is either of the EvaporativeCooler:Direct:CelDekPad or EvaporativeCooler:Direct:ResearchSpecial as opposed to the humidifies which heat the water. There is no reason to add heat to the airloop serving the computer equipment as the computer room never, ever needs heating.

However, I've run into a problem on how to control the direct evaporative coolers based on humidity. The ideal relative humidity for computer equipment is 45% - 55% and I don't see an easy way for the EvaporativeCooler:Direct to be turned on and off based on the humidity. What object would you recommend that are smart enough to control the fan, cooling coil, heating coil that is in in the airloop but is never used, and evaporative cooler to meet a temperature of 28C and relative humidity kept between 45% - 55%?

A possible alternative is to use a Humidifier:Steam:Electric object and specify the parameters in such a way that it evaporates water but generates zero heat. Then there are appropriate objects for controlling the equipment in the airloop based on temperature and humidity while not unnecessarily heating the computer equipment with a kluge parameter choice of the Humidifier:Steam:Electric object.

POST SCRIPT
Providing some data from the simulation, the humidity can get quite low if there is there is no humidifier present. These are the relative humidity values for the first day of January and the first day of July in a data center that does not possess a humidifier.

Date/Time   DATACENTER_MID_ZN_6:Zone Air Relative Humidity [%](Hourly)
01/01  01:00:00    29.96093349
01/01  02:00:00    29.95718355
01/01  03:00:00    29.9684331
01/01  04:00:00    29.96843366
01/01  05:00:00    29.98718226
01/01  06:00:00    29.97968237
01/01  07:00:00    29.98718184
01/01  08:00:00    30.01718079
01/01  09:00:00    30.04717912
01/01  10:00:00    30.07717769
01/01  11:00:00    30.0884264
01/01  12:00:00    30.06967808
01/01  13:00:00    30.03967937
01/01  14:00:00    30.0284301
01/01  15:00:00    30.02842996
01/01  16:00:00    30.02842954
01/01  17:00:00    30.0096808
01/01  18:00:00    29.99843139
01/01  19:00:00    29.99843181
01/01  20:00:00    29.99843125
01/01  21:00:00    29.99843167
01/01  22:00:00    30.01718055 ...
edit retag close merge delete

I agree that the cooling coils will remove humidity, but that humidity will stabilize at around 40%RH. This is because the cooling coil discharge temp will be at or above the room dewpoint - resulting in no further condensation or %RH change. Assuming that makes sense, and that 40%RH is OK (though perhaps not "ideal"), I don't think you actually have a problem. The alternative is to constantly spend energy adding and removing moisture.

( 2019-04-13 14:35:33 -0500 )edit

@Jim Dirkes The relative humidity does get below 40%. That number depends on the humidity outdoors and in nearby zones(leaky walls), and the surface temperature of the coil. Please see the updated post.

Also, even if the relative humidity stabilized at 40%, that is still too low for the computer equipment. I still need a way to control the moisture in the air. A method to control the evaporative media could be very useful to other people as well, such as supplementing an HVAC system with a EvaporativeCooler:Direct:CelDekPad but turning it off when the humidity gets too high.

( 2019-04-15 11:12:22 -0500 )edit

Humidifying the air adds (latent) heat to the space. Also, I do not think you would use a swamp cooler for a data-center. Are you worried about designing the system (which looks pretty easy to do), or the energy consumption?

( 2019-04-15 11:30:29 -0500 )edit

@luisclv I'm confused about the "humidifying the air adds (latent) heat to the space" sentence. Is that a bad thing?

I do agree that the swamp cooler is not ideal. The EvaporativeCooler:Direct:CelDekPad just seemed like the best object available. Are there any objects you would recommend for adding moisture to the air?

I am most worried about controlling the humidity of the air, then trying to lower the energy consumption while maintaining the humidity and temperature.

( 2019-04-15 11:40:04 -0500 )edit

@mldichter I was referring to your statement "A possible alternative is to use a Humidifier:Steam:Electric object and specify the parameters in such a way that it evaporates water but generates zero heat.". It sounds like you meant that you do not want to introduce any more sensible heat to the space. What is the system in real life? An AHU plus a humidifier? I think that your best bet is to mimic the equipment in real life with what is available in energyplus.

( 2019-04-15 11:53:52 -0500 )edit

Sort by » oldest newest most voted

I got something working. I put a splitter in the HVAC so that the coils were separate from the humidifier. One side has a fan, cooling water coil, and heating water coil. The other side has a fan and a direct evaporative cooler. Then I used some energy management objects to turn on and off the availability schedule of the evaporative cooler based on the humidity of the computer room. I try to avoid EMS objects when I can, but in this case they are working pretty well. Having the availability of the evaporative cooler lag by a timestep probably won't cause a problem anyway.

Putting the humidifier before the cooling coil didn't work since the coil condensed most of the moisture out of the air.

Putting the humidifier after the cooling coil didn't work because the air coming into the evaporative cooler was 15C so the air couldn't hold much water. The air coming out of the evaporative cooler was almost 100% relative humidity, but once it got into the data center, the relative humidity couldn't get above 45% since the data center air was 28C.

I also tried using controlling objects from some energyplus example files, but those controlling either don't directly control the direct evaporative cooler objects of just didn't work that well.

more

Your Airloop will look: Humidifier => Cooling Coil => Heating Coil => Fan

You will use SetpoinManager:MultiZone:Humidity:Minimum and SetpoinManager:MultiZone:Humidity:Maximum in the nodes after the humidifier and the cooling coil, respectively. To do this, you have to specify the humidity ratio for the spaces. Decide what is your space setpoint and calculate the humidity ratio from there. From your requirements (28C and between 45% RH to 55% RH) the minimum HR would be 0.0106 and the maximum 0.013. When looking at the results, either look at the zone humidity ratio or export both RH and dry bulb temperature.

You will also need to specify the Control Variable for the Controller:Watercoil for your cooling coil to be TemperatureAndHumidityRatio.

You cooling coil can be any type accepted in an airloop, so it does not necessarily need to be Coil:Cooling:WaterToAirHeatPump:EquationFit, it could also be a Coil:Cooling:Water or any other type your system uses.

You said that you do not need heating, but you could need to reheat the air after dehumidification to prevent overcooling (I am not sure if this is allowed by code in a data-center).

You can find an example of this arrangement in 5ZoneWaterCooled_MultizoneMinMaxRHControl of the Energyplus example files.

more

@luisclv I modified the example file you recommended and added the output:variable for humidity to my post since I couldn't fit the spreadsheet here. Did that file have all the necessary objects in it?

( 2019-04-15 17:29:40 -0500 )edit