Calculating air flow rate for UAFD system

[ypwtb]

Hello,

This web page

http://www.cbe.berkeley.edu/underfloorair/designguidelines.htm

says:

Research indicates that stratification for UFAD systems can result in overall delta T's (return-supply temperature difference) in the range of 8-11°C (15-20°F), for properly designed systems. However, these values are not what determine the air flow requirement.

Then what does? It seems that they try to explain:

The heat gain to the occupied zone and the air flow required to maintain a given comfort condition in that zone is what determines the actual air flow requirements, and consequently the overall delta T that will be developed.

If there is a maximum acceptable ceiling-temperature value and a minimum acceptable supply air temperature, then a given total cooling load (regardless of how the cooling load is distributed by elevation) seems that it is going to dictate the air flow rate.

Why wouldn't the overall room delta T and the sum of all heat gains (regardless of elevation) be the bases for calculating the air flow requirement?

Thanks,
Yittri

 

[gepman]

I am not going to pretend that I understand these underfloor systems. The company that I work for has written a white paper which includes this type of system. I will see if it is available for release and if so direct you on how to get it. Following are my guesses, not answers.

1) I think that I would get the ASHRAE design guidelines which are mentioned at the website you referenced.
2) I think the referenced website feels that a lot of the heat gain is in an "unoccupied" zone, i.e. either above where people are or in locations where people are not. Remember the idea behind this type of system is that the air flows from the floor straight up, it is NOT well mixed like a ceiling distributed system. Therefore if you have a large heat source (such as a large computer) on the floor the heat flows straight up and there are no people above the computer so you do not have to be concerned about the temperature above the computer. Think of all this heat in the same way that you think of heat gained from a plenum in a plenum return system, it doesn't affect the occupants. The heat gain in a plenum return system affects either the amount of supply air or the temperature of the supply air but of course it does not affect the total heat load.
3) I think that these systems try to minimize the amount of air and have a high proportion of exhaust air. Supposedly that is supposed to save energy. It would if the great proportion of the heat is gained in the unoccupied areas, then both the supply air volume and the supply air temperature would not have to take these heat gains into account especially if all of the air is exhausted and not returned to the coil to be cooled. I have not gone through the design of any of these systems and as one that is familiar with convential ceiling supplied systems with "complete" mixing my initial thought is that to have these low supply air volumes coming from the floor everyone is going to have frozen toes.

As I said these are my guesses. I have only heard of these systems and read a couple of white papers on them. I am somewhat skeptical but apparently they have been used in Europe in certain locations.

 

[GMcD]

Get the ASHRAE Guide to UFAD and Displacement Ventilation design and go though some of the calculation examples. The main energy saving features of the UFAD system are the higher supply air temperature (63F to 66F) allowing more range for free cooling with outdoor air, as well as the lower overall supply air rates required for a given space and cooling load.

The real key as you are realizing is that the normal cooling load calculations don't work. For UFAD and DV air systems, you have to discount much of the heat gains in the space due to stratification of warm air of heat sources. For example, and speaking in general here, you only take 15% of the lights, and 50% of the equipment and people heat gains into account as part of the "room cooling load" for calculating supply air quantities. The iterative process then requires a back-check of the stratified layer height location relative to the occupied zone, and then adjust the supply air quantity based on that, as well as humidity control in the space. In general, the UFAD systems only need to supply between 60% and 75% of the airflow compared to a conventional overhead mixing systems. This will vary depending on your local climate zone and outdoor/indoor humidity control requirements, as well as the room height for stratification.

Now, that being said, the actual cooling load on the cooling coil in the AHU will be the sum of 100% of ALL the heat gains in the room, since the return air temperature is higher than with conventional mixing air systems. You end up with a lower airflow but maybe a fatter cooling coil at the end of the day depending on dehumidification requirements.

There are two different cooling load calculations you have to do- one for the room supply air calculation, and another for the AHU cooling coil load. There are NO FROZEN TOES with these systems since the supply air temperature is above (should be) above the threshold of cool drafts on the lower limbs (important with a female occupancy). The lower limit for supply air temps for UFAD and DV systems is about 63F (17C), and the ideal range is 64F to 66F. Don't forget to take a small sensible cooling load credit for the cooler floor temperature - you have a huge radiant surface at around 67F-68F which will provide around 8-10 Btuh/sq.ft. worth of sensible cooling.

 

[gepman]

What GMcD says makes sense. I just assumed that a UFAD (UnderFloor Air Distribution) system was a DV (Displacement Ventilation) system. However the way that GMcD describes it the total heat load at the coil would still be the same as a (ceiling) mixed air system although I can see how you could save energy since both the fan energy would be less and you could have higher suction pressures on your compressors (since the supply air temperatures are higher). Since the return air temperatures are much higher than with a mixed air system I would think that there would be a much higher percentage of the time that the outside air enthalpy would be lower than the return air enthalpy and you could save energy at the economizer by bringing in outside air. Reduced contaminants in the air (higher IAQ) is supposed to be another benefit of DV systems.

 

[GMcD]

Gepman- gotta be a bit careful about assuming that UFAD is the same as DV systems. While they share some characteristics (stratified air layer) and similar supply air temperatures, the UFAD systems are a "low level mixing" system using high velocity air supply outlets that mix the air in the occupied zone to get relatively uniform air temperatures in the floor to 6 foot level.

DV systems use low velocity air supply at low levels (less than 70 fpm supply air velocity) to create a pool of cooler air that warm objects "draw" towards themselves due to the thermal plumes off said warm object. DV is almost a self-regulating system due to this effect, but relies on a fairly undisturbed environment to work well - sedentary offices, low people movement activities. There is a much more distinct temperature gradient from floor to ceiling with a DV system compared to a UFAD system. So a DV air supply system can be supplied by an underfloor plenum, if required, but DV can also be supplied by low sidewall air outlets as well. A DV system is another flavour of a UFAD system.

UFAD systems come in many flavours: pressurized plenum; neutral plenum with powered air terminals; ducted air supply to air terminals; semi-ducted systems with floor plenum dividers, etc.

Hopefully Ytwtb can google "UFAD system design" and get a lot more information, there have been some good articles in HPAC Magazine and other sources on design issues to be dealt with.

 

[atlas06]

GMcD
I do have the ASHRAE guideline you are talking about, but we are seeing these UFAD systems where air-side economizers do not apply (high end high rise new buildings), and if you add the terrorism protection that all OA must be above the 4th level (in the US), there goes the air-side economizer and the whole point of UFAD.

You mention that only 15% of lights and 50% of people and equipment should be accounted for in Load calculations. I am sorry, but I fail to see your point, if indeed that is the case, then the comparison of UFAD Vs Standard VAV at same airflows makes sense, but I am not convinced of the reduction in cooling load calculations.

Can you elaborate on the reduction in cooling load and direct us to some scientific published data substantiating the use of 15% lights and 50% equipment/people for load calcs?

Thank you very much for all your inputs into this forum, you are one the most contributing guys here.

 

[GMcD]

Atlaso6: just Google Displacement Ventilation and "underfloor air design" for all the scientific back-up. It's not "my point", it's basic physics. The ASHRAE Guide for UFAD and DV system design has this calculation method in there to pro-rate the heat from lights and heat producing elements in the room.

With an air supply at floor level, the heat produced by lights aren't part of the room air conditioning load because the heat produced by the lights goes directly into the ceiling exhaust/return plenum and does not produce any cooling load down in the occupied space. Heat from fluorescent lights is mainly convective heat off the ballasts, and there is very little radiant heat gain from the lights to the occupied space. In a full mixed air system that you get with overhead cool air supply, you have to account for that heat off the lights as part of the room cooling load because you have to mix that heat at the ceiling level with the air supply to get uniform air temperature in the occupied space - that's the basic overhead air supply system method- getting fully mixed uniform "air temperature" conditions in the occupied space. Ditto with the thermal plumes from heat producing sources with a UFAD system - part of the heat gain from equipment and people is carried off in the thermal plume and goes up into the ceiling return/exhaust, and does not become part of the "occupied zone" cooling load. With a UFAD system You ARE NOT creating a uniform air temperature in the occupied space - you are creating a stratified air temperature in the space such that the "comfort temperature" is maintained up to the 7 foot elevation, and then it gets warmer as you go higher in the room.

I don't understand how an air side economizer WOULDN'T apply to a "high end high rise new building". Why can't you use as much fresh air for free cooling as you can to minimize energy? My design philiosophy is to beat up the Architects to get as high a performance envelope and facade as possible to minimize the transient thermal loads in the space, and then you can start to use semi-passive thermal comfort systems like radiant heating and cooling with dedicated outdoor air systems (DOAS). Now the DOAS can deal with your terrorism and outdoor air control and treatment, while hydronic radiant systems can deal with most of the comfort control in the room.

 

[marcoh]

I have jsut started work on my first UFAD displacement system, and ASHRAE 'System performance evaluation and design guidelines for displacement ventilation' is a very handy guide to start.

As per any building the biggest challenge is the architect who wants 4m high windows facing due west with a 'clear' look!

 

[GMcD]

Marcoh: Fricken architects - push hard on the question "why?" they "need" this amount of west facing glass. Even with the worlds best HVAC system (picking "best" is relative to the designer) there WILL be discomfort and poor indoor environmental quality along that facade no matter what you do.

What you need to do is show a cost-benefit analysis with say 2M high glass vs the 4m High glass and show the savings in HVAC capital costs and operating costs, as well as the reduced glare, reduced radiant heat off the warm glass as the sun hits it, and the how view won't be there when everone closes the blinds to keep glare off their computer screens anyway.

 

[marcoh]

The 'why' is easy to answer, it is vital element of the 'vision' for the building! But we are going through all the exercises, heat gains, comfort factors, glare issues, costings, energy etc to show the effect and slowly the 'vision' is becoming more realistic.

It has been an eye-opener to see how glazing technology is progressing, especially some of the German/Swedish manufacturers claiming amazing figures (eg U<1, SC<0.2)

 

[GMcD]

Marcoh- You're on the right track. The Architect can have his 4M high glass, BUT, it has to meet "this" minimum performance (and you wave aside your super-engineer cape) and pull out the Visionwall/Geilinger glazing catalogue to tell him that the glass has to be at least a 4-element unit, maybe even with Argon or Krypton fill to get that R-10+ performance. Still won't do much for the low angle sun solar loads which will be the killer for glare and solar gain, even with high performance solar gain numbers.

 

[gepman]

marcoh
you can easily model all of the buiding elements and change different properties and make parametric runs with eQUEST(you may be using it now). You can model the glass and glazing systems with Window5 and import the properties into eQUEST. I think that you can make some type of modifications (although you need to know the inner workings) to eQUEST to model UAFD and DV systems. eQUEST is available for free at

http://www.doe2.com/equest/

and Window 5 is available for free at

http://windows.lbl.gov/software/window/window.html

Make certain any window properties claimed by the manufacturer have been tested to NFRC. Window 5 usually has all of the properties as a function of solar angle which most manufacturers do NOT list.

You can take courses on using eQUEST for free in California (since it is sponsored by the state electrical utilities). It is a powerful design tool and also can determine compliance with Califoria T24 energy code which is much more stringent than ASHRAE 90.1.

Depending on the size and shape of the building you can have preliminary results in 2 to 4 hours of work with eQUEST and Window 5.

 

[atlas06]

GMcD
I see all the hooplah about the space not seeing the load, but the coils do, I just see this whole thing as experimental. But, tehoritically, I tend to accept the point.
I do like the DOAS with radiant beams, but it's a hard in the US where Humidity IS a problem (keep in mind that we have 60% of teh world's lawyers).

Have you tried this UFAD for a slab-on-grade application? GSA here recommends against it.

Thanks

 

[GMcD]

Atlas06: There are a number of UFAD systems on ground floors (slab on grade) buildings down in the Seattle region (Microsoft buildings mainly) and they all work fine. Like any integrated system, the first thing to do is get the insulation under the slab to decouple the ground floor slab from the ground conduction, and then do a semi-ducted in-floor distribution duct system so that air terminals are not more than 50-60 feet from a supply duct outlet in the floor for the big floor plates.

I'm not a fan of passive chilled beams since those are designed to run right at the ragged edge of condensing in order to get a cool enough boundary layer of air to start departing and "falling" down from the chilled beam, and for the "active" chilled beams, where the supply air is pushed through them (the chilled beams just act like a re-cool coil in that case anyway), the fundamental issue is that it's still an overhead "mixing" application where the polluted stratified air at the ceiling level is mixed and recirculated back into the occupied zone. Sure, there is a little bit of radiant cooling off the active chilled beams, but they have such a small surface area that it's not a dominant cooling element anyway.

Gimme low level air supply, and stratify the polluted air to the ceiling exhaust stream and I'm a happier guy. 100% outdoor air DOAS systems with heat/energy recovery and use hydronic radiant systems for most of the space temperature/comfort control.

 

[gepman]

Earlier I promised a white paper on DV. You can find one at

http://www.energydesignresources.com/resource/199/

There is also a light of other good energy efficiency resources there also.

I also thought that someone might be interested in the following link since DOAS was mentioned.

http://doas-radiant.psu.edu/doas.html

 

[atlas06]

GMcD
Sorry for being the devil's advocate but the problem I see with slab-on-grade application for UFAD does not come from the heat exchange with the slab but from the quality of the concrete slab construction. I have seen several slabs (in parking garages mainly) that have so many cracks in them that I do not trust that the floor will not present cracks over the years, and thus making the air contaminated by seapage of ungerground water, vermine of all sorts, atc..

It may be OK for teh first few years, but ina long term, serious leakage problems will be seen in my opinion. The hard part is that cracks in a raised floor can go on for years without ever finding out.

 

[GMcD]

Atlas06: Typically the slab on grade insulation is placed on the sub-grade, then poly vapour barrier is placed over that, then the concrete gets poured on that. Sure, there are shrinkage cracks in the concrete but seepage of ground water, radon gas, etc, are generally prevented by that construction detailing. If you've read through the LBL UFAD material, they also recommend applying a concrete sealer on the exposed concrete in the floor plenum. Parking garage slabs are very different from occupied building slabs, and one would think more care would be taken to pour and screed the occupied building slabs properly, compared to a parkade slab.

C'mon, we're engineers, our job is to think about the "what-if scenarios" and design the means to mitigate that risk. If the building is in a climate zone where ground level bugs are a problem, design the footings and slab perimeter detailing to mitigate it. If there is a high water table, then there ought to be a lot of good sub-grade drainage being used as well, to prevent water coming up through the slab.

I don't personally like UFAD systems as a "go-to" system due to the housekeeping issues of keeping the pleneum clean, spills into the plenum, potential pipe leaks into the plenum, and the sheer amount of frustration trying to get the Architect and General Contractor to police up all the plenum air leakage paths. BUT, they are a toolkit system that we need to be aware of, and in spite of the potential risks, there are other advantages in certain applications that outweigh those risks, and a UFAD system may be a valid application for a particular project.