NEW (?) eQUEST DOAS workaround using plenum spaces

11 posts / 0 new
Last post

I doubt I'm the first to think of this but I don't recall it being discussed on the List. Simply put, you can model a DOAS system by assigning it to the plenum spaces above occupied spaces, and putting all OA loads in the plenum zones.
I propose the following workaround for modeling dedicated outdoor air systems (DOAS) in eQUEST/DOE-2.2:

- Model all shells/floors with plenum spaces if they are served by a DOAS

- Change the plenum zones ZONE: TYPE to "Conditioned"

- Leave the plenum zone COOL-TEMP-SCH and HEAT-TEMP-SCH as "undefined" (no zone control over the DOAS operation)

- Pick DESIGN-COOL-T and DESIGN-HEAT-T values for the plenum zones that match the likely DOAS supply air temperature range

- Create the DOAS as a new SYSTEM (TYPE = [whatever is appropriate])

- Assign the DOAS to the plenum zones above the conditioned zones that it actually serves

- Assign outdoor air only to the plenum zones, representing the outdoor air requirement of the conditioned zones below

- Create/modify internal walls between plenum zones and zones below, making them "Air walls"

This method has the following advantages:

- No exceptional calcs or separate models needed (but see "Conditioned area" comment below)

- All DOAS systems can be modeled as designed, within normal eQUEST constraints

- A single DOAS system can supply ventilation air to multiple zones/systems

- DOAS system operation and ventilation supply can be independent of the other systems serving the zones

- Return air temperature to the DOAS closely follows that of the occupied zone, for accurate heat recovery simulation

- Zonal systems "see" the space load from ventilation air that is not supplied at a "zone-neutral" temperature

- DCV can be modeled, but people need to be assigned to the plenum spaces instead of the occupied spaces below (which might also be a disadvantage of this method)

- No ventilation load changes need to be made to create the LEED Baseline from the Proposed design model, but the plenum zones need to be assigned to the baseline systems that serve the zones below.

...but has these disadvantages or does not solve these problems:

- Conditioned floor area reported (such as in LS-C) is incorrect. Subtract the area served by the DOAS (as reported on SV-A).

- Latent heat transfer is not simulated across interior walls. Ideally, latent loads should be assigned to the plenum spaces to represent the portion of non-ventilation latent loads handled by the DOAS.

- Plenum space FLOOR-WEIGHT may need to be adjusted for realistic lag time and peak load damping (I haven't thought this out yet)

Please discuss!

~Bill

William Bishop, PE, BEMP, BEAP, LEED AP

Bill Bishop's picture
Offline
Joined: 2012-02-25
Reputation: 7

I?ve never used this method, but at first glance it looks like reasonable approach to me. It seems like a key assumption is that using an air wall between the space and plenum acts like air mixing between the spaces, so loads in each space are seen by the other. That is an assumption I would run some hourly reports and tests to verify. Have you already done that?

___________________________________________
Jeremy McClanathan, P.E., BEMP, HFDP, LEED? AP

Jeremy McClanathan's picture
Joined: 2011-09-30
Reputation: 200

Yes, air walls are a key component of this workaround. I have used hourly reports to verify that plenum zone temperature tracks the occupied zone temperature below fairly well, as shown in this screen shot of a perimeter zone (2nd column is plenum zone; 3rd is the zone below):

[cid:image002.jpg at 01CEDAFD.31FA9590]

Bill Bishop's picture
Offline
Joined: 2012-02-25
Reputation: 7

I'd say that is a pretty clever approach. Would you just add the geometry as a normal plenum zone with exterior/interior walls and such or would you just have them exist as plenum zones with only one heat transfer surface in common with each the zone(s) its serving? Wouldn't you want DESIGN-COOL-T AND DESIGN-HEAT-T to be the same as the zone it is serving? Then you could set the u value of the air wall high so that any cooling or heat that the DOAS provides would then be effectively transferred into the occupied spaces. I think in that scenario it might act as a quasi supply plenum? Not sure without actually trying it. Just my first thoughts, maybe you have already considered all that and the way you described works better.

Brendan Hall, PE, LEED AP BD+C

Hall, Brendan's picture
Offline
Joined: 2011-09-30
Reputation: 1

You make no changes to the plenum space geometry as it is normally created via the eQUEST wizard, except to make sure there is a ceiling, and change the ceiling surface TYPE to "Air". I used the "Default Air Wall Construction" which has a U-value of 2.7 Btu/ft2-hr. I'm not sure about DESIGN-COOL-T and DESIGN-HEAT-T. I think you would want them to represent the DOAS SAT, so that the system(s) serving the occupied zones below can be autosized accordingly (as described in the DOE-2 help menu). I just noticed that you have to use "Adjust Loads" for the zone SIZING-OPTION for this effect to be included.

~Bill

Bill Bishop's picture
Offline
Joined: 2012-02-25
Reputation: 7

Correction - ceiling surfaces are created for the conditioned spaces below the plenum spaces, so the change to "air wall" is made for the space below, not in the plenum space. You do, however, need to change the plenum zones to "conditioned" in order to assign OA to them. I'm also thinking it might make sense to assign thermostat schedules to one or more DOAS plenum zones, to control whether the DOAS is in heating or cooling mode. You could use schedules with lower heating, higher cooling temps and/or larger throttling range and/or different t-stat type, in order to avoid increasing the reported unmet load hours.

~Bill

Bill Bishop's picture
Offline
Joined: 2012-02-25
Reputation: 7

My past efforts at approximating DOAS and MAU (makeup air unit) setups in eQuest have sometimes paralleled the approach Bill is describing.

Some extra thoughts/cautions from my experience:

- Common scenario: A DOAS pre-conditions or does not condition ventilation air beyond heat recovery before supplying to the main conditioning units, which mix that OA with the return airstream and heats/cools as needed to arrive at the intended supply air conditions.

- An important point is this approach "circumvents" typical supply air conditioning means by dumping the associated ventilation loads (after any DOAS-borne recovery & pre-conditioning) "directly" into the spaces. This may have effects of:

o Overstating the actual ventilation loads to be dealt with(where return/relief air would otherwise mitigate/eliminate)

o "Moving" where ventilation loads are dealt with from a central system to terminal (zonal) units,

o Causing the conditioned zones to call for heating/cooling where they might otherwise call for the opposite, which may throw off expected behaviors, terminal unit sizing, and in some cases may have effects on how well the main conditioning units handle the situation (i.e. unmet hours)

o Example for illustration: Typical office building with a DOAS, furnace heated air handler which delivers 55 degree SAT, and electric reheat VAV boxes. A core office zone might call for cooling from the main system in the winter (internal loads dominant), but the DOAS-supplied (and air wall transferred) ventilation air loads may instead turn the space into a refrigerator for the hour, so the zone thermostat would calling for heating instead... suddenly your whole building is dealing with all your wintertime ventilation heating with electricity instead of gas as anticipated.

You can try to compensate for some of these effects by moving or mirroring your main air handling conditioning sources/capacities into the DOAS system, but at some point you're facing a potential documentation nightmare for LEED or similar...

I don't mean to be a negative Nancy - this is the clearest procedural write-up I've run into for a "dummy zone" approach to DOAS behavior (more on this in the archives).

The other major alternatives to a "dummy zone" workaround are, in brief:

* Opt to not model the DOAS system as a separate system for the simulation, and instead define its components within the served systems. Put another way: Add heat recovery recovery/pre-conditioning/fan power/etc in proportion to the systems the DOAS serves.

* Look to software other than eQuest if none of these approximations are acceptable.

Best regards,

~Nick

[cid:489575314 at 22072009-0ABB]

NICK CATON, P.E.

Nick-Caton's picture
Offline
Joined: 2011-09-30
Reputation: 605

Hats off to you, Bill. It looks like a clever solution.

I have noticed that for some HVAC systems the wizard creates one big plenum over all the spaces and for other systems it creates individual plenums over each space below. If one big plenum was created, then you would have additional steps of breaking up the plenum and adding interior walls.

Keith Swartz, PE, BEMP, LEED AP

Keith Swartz's picture
Offline
Joined: 2011-09-30
Reputation: 0

Keith,

Yes, if you know you're going to get into this level of detail in a model, it makes sense to select "System per Zone" in the wizard so that individual plenum spaces are created above each space. It would be a nightmare to create them after-the-fact.

Thanks,

Bill

Bill Bishop's picture
Offline
Joined: 2012-02-25
Reputation: 7

A trick to get eQuest to create a plenum per zone with "System per Floor" is to temporarily move a corner of a space that is shared with another space, leaving a small gap between the spaces. Then, when you get to the detailed mode, move the corner back.

___________________________________________
Jeremy McClanathan, P.E., BEMP, HFDP, LEED? AP

Jeremy McClanathan's picture
Joined: 2011-09-30
Reputation: 200

Nick,

Thanks for your valuable feedback! I agree that the plenum workaround I described is not ideal for all DOAS systems. In the past, I have used the option you describe below, where you don't model a separate DOAS - you just add heat recovery, additional fan power etc. to the other systems as appropriate. I came up with this workaround for a system configuration that I am starting to see frequently - a packaged DOAS (w/CHW or DX cooling and/or heat pump) supplying OA either directly to zones or to zonal systems (usually water-to-air heat pumps). Regarding your specific concerns:

* Overstating ventilation loads - Not sure I understand. The method I described takes the design OA cfm and preconditions it at the DOAS as designed before sending the ventilation "somewhere" (in this case, above the occupied zones), so the load is what it is. Maybe you are just restating your second concern, which is...
* Moving the ventilation loads - I agree. You could potentially shift a gas load to an electric heating load, or alter the diversity of heating/cooling modes in your terminal systems. I would think this would be more of a problem if the air-side has a high OA%, and/or if the DOAS supplies air at a non-zone-neutral temperature. For the first model I am trying this on, the DOAS SAT is close to room temperature, largely due to heat recovery, and the terminal units are water-to-air heat pumps which use electricity for both heating and cooling and reject/absorb heat from the same heat pump loop.
* Unit sizing and unmet hours - Adjusting the DESIGN-COOL-T and DESIGN-HEAT-T should help with baseline autosizing. If using eQUEST for design you would certainly use design-case space loads. Since the selection is usually larger than the peak load (manufacturers only offer incremental sizes except for custom equipment) unmet hours shouldn't be much of a problem. Again, high OA% or non-zone-neutral OA is a caveat.
* I wouldn't consider this another "dummy zone" workaround. Indeed, a big advantage is that you don't create any virtual (not physically part of the building) zones or virtual/estimated loads.

Thanks again,

Bill

Bill Bishop's picture
Offline
Joined: 2012-02-25
Reputation: 7