"if the air in duct is the same air as the air being supplied to the space,
there is no net effect"? What do you mean by net effect?
The duct heat loss/gain depends on the difference of the temperature of the air
in the duct and the temperature of the air surrounding the outside of the duct.
It is highly unlikely that these temperature will be the same.
See page 14 to 17 "Thermal Analysis"
http://bepan.info/class-notes/e6_-_loads-ducts-pipes
AES-Ductwork-System-Design
When the cooling DT (Room-Suppy Temp) is 20 degs, every degree temp rise from the cooling coil affects the supply air by 5%. For a draw-thru system with a supply static of 8" to 9" the heat gain across the fan, the temp rise can be significant. The rule of thumb in the old days, when no one could be bothered doing these calculations manually, the temp-rise was assumed to be = 0.5 x SP.
The velocity upstream of the terminal box from the AHU through the unconditioned shafts of high-rise bldgs is typically about 4,000 fpm which results in low heat gain due to the high velocity. The terminal box and the low velocity ductwork is above conditioned space (typically RA plenum). If the ductwork is insulated on both sides of the TB, the rule of thumb for a temp rise through the ducts is about 1 - 2 degs which requires about a 10% increase in supply air.
Before anyone replies to this message saying that it has wandered off the point from the issue being discussed, I will admit the reply is an excuse to for mentioning the website. I am thinking of closing it down if there is no interest from others in developing building energy analysis course material from a full design perspective and not just using computer program. The emphasis would be on teaching Energy Efficient Building Design with project case studies. It does not quite fit in with BEMBook The objectives are in the Introduction http://bepan.info/introduction
Varkie
I think Mike's general point is fair and isn't to be taken in absolutes.
When a ducted return air stream is passing through a conditioned space
or plenums/chases in between other conditioned spaces, it's probably a
safe bet to claim any duct gains/losses are *negligible* (not
nonexistent), particularly if said return ducts are insulated, due to
the relatively small delta-T.
On the other hand, winter losses and summertime gains at extreme
temperatures may be considerable if any ductwork is routed exterior to
the building, even if insulated.
Xiaoyang, if you wish to calculate it as a percentage at design
conditions: ( U x A x deltaT )DUCT / (total building heating or
cooling load) x 100
If you wanted to model such exterior duct heat losses/gains in eQuest...
I haven't considered that before - but depending on your system setup,
it might be as simple as a small exterior wall surface with appropriate
U-value and area equal to the duct surface area... in the event the
solar loads incident on the duct might be significant, make a point to
correctly locate the surface in 3D (over the roof, along a wall, etc.)
to avoid any incorrect self-shading from other surfaces.
If anyone has delved into this and knows a better/simpler way please do
share =).
I'm not currently an educator, so perhaps I'm not the intended audience,
but as an MEP designer actively practicing energy modeling in industry
(outside of academia) I've stumbled into and found the BEPAN site to be
very handy multiple times. Anyone who hasn't perused the site might
want to dig around a bit... there's quite a few tools/spreadsheet
examples provided which I've found interesting-to-useful, though finding
any one thing can be tricky as it's quite a lot of material!
I hope the site doesn't go down, or else if it should have to you'll
give us a fair warning to scramble and make copies of useful
references/tools =). It seems to exist solely as a good-faith
contribution to helping others learn from your past experiences, and for
that I hope you know I for one very much appreciate your efforts!
For a supply duct passing through a plenum return ceiling during the cooling
season, the temperature difference between the supply air and return air can be
as much as 25 degrees.
During the heating season, this difference could be as much as 30 degrees during
morning warm up.
This calculator is really meant for exterior applications as well as determining if condensation will occur (and you have to use their preset products for insulation thickness & k values), but it can be a quick reference for us looking to see what an avg heat loss might be in a duct system. (I?m sure there?s a reference to the ASHRAE handbooks, but I?m drawing a blank right now?)
What I am unsure of in eQuest is how it uses Supply Duct UA and Duct Delta T. If heat is lost in the duct, is it gained in the ceiling space or plenum?? If that is the case, then the arguments below that the overall net effect is negligible is a good one. However, fan energy will increase since the delta T of the supply air to room air decreases.
In addition to significant delta Ts, please also consider the additional airflow required to cool/heat a space when the supply air delta T (between the supply air and the conditioned space) is lowered due to duct losses.
Example: Say the cooling load requires 100 cfm to cool a space assuming a supply air temp of 55 and a perfectly insulated duct (that is, a duct with no change in temp as it leaves the AHU and gets to the space). The room is 70 degrees, so which makes the sensible cooling in the space about 1630 btu.
Now take this same room, and assume the duct gains 5 degrees between the AHU and the space (note, I am not talking about heat gain from fan energy here). Now the air being supplied to the space is 60 degrees, and to maintain the same sensible cooling capacity, the airflow must be 150 cfm, a 50% increase (!).
This heat is not lost to the space being cooled (most likely). The idea of *neglible* loss at the room level does not apply.
At the system level, the idea of *neglible* energy loss may or may not apply. CLEARLY there is additional fan energy used. The heat that is lost from the duct to somewhere, may be recovered to a large degree (in the case of a 100% return air system, where the primary area of energy loss COULD be additional heat gained in the plenum due to increased delta t between the plenum and the OA temp) or it may be lost entirely (in the case of a 100% OA system, which is obviously already a big energy user and this simply adds to the problem). In the case of significant relief of the air the heat is lost to, there is also a large impact on cooling and heating energy as you are moving more air, and trying to maintain the same setpoints at the system level.
The notion that there is no net effect is really only true if the heat is being lost/gained directly at the location of the temperature sensor/tstat.
(Most buildings these days have appropriate insulation on the ductwork which minimizes this effect, but it wasn?t that long ago when this wasn?t the case)
Thanks very much for the clarifications! I?m learning a thing or two, but I suspect my choice of words might be getting in the way of what I was trying to convey?
Please consider and affirm or correct this statement:
If there is rarely a case where return duct plenum losses/gains are negligible, then it follows all energy models should explicitly account for these losses/gains to those spaces.
A specific follow-through would be: If an eQuest model does not model conductive heat gains/losses in the return air path, does that make the eQuest model invalid?
Trying to pull this discussion back into daily practice with eQuest? I am still fuzzy on exactly how and to what extent eQuest/DOE2 does model such internal supply and return duct losses/gains (outside of light load fractions)? Can anyone in the know clarify this query?
The original post as well as follow ups from Paul and myself discuss loses that occur in supply and return ducts, not limiting the discussion to return ducts nor limiting it to plenum loses, for that matter.
Equest has inputs for duct loses, in several different formats/locations. In my experience, it has been difficult to accurately model these loses and obtain logical results.
I believe Mike?s point was ?if the air in the duct is the same temp as the air the duct is passing through, the effect is negligible?. I certainly agree with that, and the classic example of that scenario is a return duct (carrying room temperature air) passing through conditioned space or plenums (assuming the temp of the plenum is near the temp of the space).
From a practical standpoint, unless you are working on an existing building with outdated construction standards, or trying to model a specific energy conservation measure that analyzes the impact of duct insulation (or potentially changing supply air temps), there are few compelling reasons to account for these loses (from an energy use standpoint).
On the other hand, if you are trying to run load calcs for equipment sizing, you may want to account for them. Again, equest has inputs under ?duct loses? to do so. It sort of sounds like this is what the original poster is asking for. (original poster: try to calculate the ducts UA value and use that).
My experience in the real world has been that the temperature in the plenum is
about 4-6 degrees warmer than the space temperature. Don't forget the heat
introduced by the lighting ballasts and the lamps themselves. I have performed
some stratification studies and the temperature of a space at ceiling level can
often be 4 degrees warmer that at the standard 5' thermostat elevation. I was
recently commissioning a project and noticed that the T-8 lamps are very warm to
the touch when they have been on for some time. T-12's run cooler. The rule
of thumb that I have used is that 30% of the lighting wattage is radiated into
the plenum. It depends on the type of lighting fixture.
On a ducted return system, the plenum is much hotter than the space temperature,
as much as 90 degrees on a hot day with a space temperature of 72 degrees. As I
work with automation systems quite a bit, it is not unusual for the discharge
air to rise 6 degrees between the discharge of the fan unit and the discharge of
a VAV box. And this is with a duct insulated with 2" of fiber. This is not a
negligible heat gain, especially when working with large air handlers.
Nick,
The website consists of my notes for teaching at IIT and case studies based on my experience at SOM. It was a hasty part-time effort. The website is paid for until December 18.
Some of the eQUEST models will generate 1000s of pages of Warning & Caution messages. I have never used eQUEST on a real project. The DOE21E models (that I might add later) are based on actual projects and might be more useful. See http://bepan.info/design 5 - Energy-Savings-Options
The Excel programs were created to solve specific project problems and they are not generic programs (could be developed into generic) that could be used on other projects.
I will continue adding the stuff lying around in my computer (from all the junk created when I was working at SOM and teaching at IIT) to the website until December 18. I still have to add my notes on Electrical, Lighting, Plumbing, & Fire-Protection.
Those with an engineering degree and a few years AE design experience do not need the information in the website. It is intended for teaching architectural students.
Varkie
Dear xiaoyang shi
It Depends on the temperature differentials between air in duct and space and insulation thickness
But more importantly , if the air in duct is the same air as the air being supplied to the space there is no net effect
Thank you, Sincerely,
Michael J. McArdle , P.E.
Michael:
"if the air in duct is the same air as the air being supplied to the space,
there is no net effect"? What do you mean by net effect?
The duct heat loss/gain depends on the difference of the temperature of the air
in the duct and the temperature of the air surrounding the outside of the duct.
It is highly unlikely that these temperature will be the same.
Paul Diglio
See page 14 to 17 "Thermal Analysis"
http://bepan.info/class-notes/e6_-_loads-ducts-pipes
AES-Ductwork-System-Design
When the cooling DT (Room-Suppy Temp) is 20 degs, every degree temp rise from the cooling coil affects the supply air by 5%. For a draw-thru system with a supply static of 8" to 9" the heat gain across the fan, the temp rise can be significant. The rule of thumb in the old days, when no one could be bothered doing these calculations manually, the temp-rise was assumed to be = 0.5 x SP.
The velocity upstream of the terminal box from the AHU through the unconditioned shafts of high-rise bldgs is typically about 4,000 fpm which results in low heat gain due to the high velocity. The terminal box and the low velocity ductwork is above conditioned space (typically RA plenum). If the ductwork is insulated on both sides of the TB, the rule of thumb for a temp rise through the ducts is about 1 - 2 degs which requires about a 10% increase in supply air.
Before anyone replies to this message saying that it has wandered off the point from the issue being discussed, I will admit the reply is an excuse to for mentioning the website. I am thinking of closing it down if there is no interest from others in developing building energy analysis course material from a full design perspective and not just using computer program. The emphasis would be on teaching Energy Efficient Building Design with project case studies. It does not quite fit in with BEMBook The objectives are in the Introduction http://bepan.info/introduction
Varkie
I think Mike's general point is fair and isn't to be taken in absolutes.
When a ducted return air stream is passing through a conditioned space
or plenums/chases in between other conditioned spaces, it's probably a
safe bet to claim any duct gains/losses are *negligible* (not
nonexistent), particularly if said return ducts are insulated, due to
the relatively small delta-T.
On the other hand, winter losses and summertime gains at extreme
temperatures may be considerable if any ductwork is routed exterior to
the building, even if insulated.
Xiaoyang, if you wish to calculate it as a percentage at design
conditions: ( U x A x deltaT )DUCT / (total building heating or
cooling load) x 100
If you wanted to model such exterior duct heat losses/gains in eQuest...
I haven't considered that before - but depending on your system setup,
it might be as simple as a small exterior wall surface with appropriate
U-value and area equal to the duct surface area... in the event the
solar loads incident on the duct might be significant, make a point to
correctly locate the surface in 3D (over the roof, along a wall, etc.)
to avoid any incorrect self-shading from other surfaces.
If anyone has delved into this and knows a better/simpler way please do
share =).
~Nick
NICK CATON, E.I.T.
Varkie,
I've found your website fascinating, personally.
I'm not currently an educator, so perhaps I'm not the intended audience,
but as an MEP designer actively practicing energy modeling in industry
(outside of academia) I've stumbled into and found the BEPAN site to be
very handy multiple times. Anyone who hasn't perused the site might
want to dig around a bit... there's quite a few tools/spreadsheet
examples provided which I've found interesting-to-useful, though finding
any one thing can be tricky as it's quite a lot of material!
I hope the site doesn't go down, or else if it should have to you'll
give us a fair warning to scramble and make copies of useful
references/tools =). It seems to exist solely as a good-faith
contribution to helping others learn from your past experiences, and for
that I hope you know I for one very much appreciate your efforts!
Thanks,
~Nick
NICK CATON, E.I.T.
Nick:
For a supply duct passing through a plenum return ceiling during the cooling
season, the temperature difference between the supply air and return air can be
as much as 25 degrees.
During the heating season, this difference could be as much as 30 degrees during
morning warm up.
Paul Diglio
If you?re looking to see what change in temp you have based on a length of duct, insulation, etc, check out:
http://www.mcgillairflow.com/textDocs/techTools/converted/thermalData1.php
This calculator is really meant for exterior applications as well as determining if condensation will occur (and you have to use their preset products for insulation thickness & k values), but it can be a quick reference for us looking to see what an avg heat loss might be in a duct system. (I?m sure there?s a reference to the ASHRAE handbooks, but I?m drawing a blank right now?)
What I am unsure of in eQuest is how it uses Supply Duct UA and Duct Delta T. If heat is lost in the duct, is it gained in the ceiling space or plenum?? If that is the case, then the arguments below that the overall net effect is negligible is a good one. However, fan energy will increase since the delta T of the supply air to room air decreases.
Andy Phelps, PE, LEED AP
In addition to significant delta Ts, please also consider the additional airflow required to cool/heat a space when the supply air delta T (between the supply air and the conditioned space) is lowered due to duct losses.
Example: Say the cooling load requires 100 cfm to cool a space assuming a supply air temp of 55 and a perfectly insulated duct (that is, a duct with no change in temp as it leaves the AHU and gets to the space). The room is 70 degrees, so which makes the sensible cooling in the space about 1630 btu.
Now take this same room, and assume the duct gains 5 degrees between the AHU and the space (note, I am not talking about heat gain from fan energy here). Now the air being supplied to the space is 60 degrees, and to maintain the same sensible cooling capacity, the airflow must be 150 cfm, a 50% increase (!).
This heat is not lost to the space being cooled (most likely). The idea of *neglible* loss at the room level does not apply.
At the system level, the idea of *neglible* energy loss may or may not apply. CLEARLY there is additional fan energy used. The heat that is lost from the duct to somewhere, may be recovered to a large degree (in the case of a 100% return air system, where the primary area of energy loss COULD be additional heat gained in the plenum due to increased delta t between the plenum and the OA temp) or it may be lost entirely (in the case of a 100% OA system, which is obviously already a big energy user and this simply adds to the problem). In the case of significant relief of the air the heat is lost to, there is also a large impact on cooling and heating energy as you are moving more air, and trying to maintain the same setpoints at the system level.
The notion that there is no net effect is really only true if the heat is being lost/gained directly at the location of the temperature sensor/tstat.
(Most buildings these days have appropriate insulation on the ductwork which minimizes this effect, but it wasn?t that long ago when this wasn?t the case)
Paul/John,
Thanks very much for the clarifications! I?m learning a thing or two, but I suspect my choice of words might be getting in the way of what I was trying to convey?
Please consider and affirm or correct this statement:
If there is rarely a case where return duct plenum losses/gains are negligible, then it follows all energy models should explicitly account for these losses/gains to those spaces.
A specific follow-through would be: If an eQuest model does not model conductive heat gains/losses in the return air path, does that make the eQuest model invalid?
Trying to pull this discussion back into daily practice with eQuest? I am still fuzzy on exactly how and to what extent eQuest/DOE2 does model such internal supply and return duct losses/gains (outside of light load fractions)? Can anyone in the know clarify this query?
~Nick
NICK CATON, E.I.T.
Nick,
The original post as well as follow ups from Paul and myself discuss loses that occur in supply and return ducts, not limiting the discussion to return ducts nor limiting it to plenum loses, for that matter.
Equest has inputs for duct loses, in several different formats/locations. In my experience, it has been difficult to accurately model these loses and obtain logical results.
I believe Mike?s point was ?if the air in the duct is the same temp as the air the duct is passing through, the effect is negligible?. I certainly agree with that, and the classic example of that scenario is a return duct (carrying room temperature air) passing through conditioned space or plenums (assuming the temp of the plenum is near the temp of the space).
From a practical standpoint, unless you are working on an existing building with outdated construction standards, or trying to model a specific energy conservation measure that analyzes the impact of duct insulation (or potentially changing supply air temps), there are few compelling reasons to account for these loses (from an energy use standpoint).
On the other hand, if you are trying to run load calcs for equipment sizing, you may want to account for them. Again, equest has inputs under ?duct loses? to do so. It sort of sounds like this is what the original poster is asking for. (original poster: try to calculate the ducts UA value and use that).
Thanks John, I think that sums it up nicely ? we?re very much on the same page now =).
My latent concerns regarding return duct inputs are resolved as well? I seem to have glazed over the right half of the basic system tab!
~Nick
NICK CATON, E.I.T.
Nick:
My experience in the real world has been that the temperature in the plenum is
about 4-6 degrees warmer than the space temperature. Don't forget the heat
introduced by the lighting ballasts and the lamps themselves. I have performed
some stratification studies and the temperature of a space at ceiling level can
often be 4 degrees warmer that at the standard 5' thermostat elevation. I was
recently commissioning a project and noticed that the T-8 lamps are very warm to
the touch when they have been on for some time. T-12's run cooler. The rule
of thumb that I have used is that 30% of the lighting wattage is radiated into
the plenum. It depends on the type of lighting fixture.
On a ducted return system, the plenum is much hotter than the space temperature,
as much as 90 degrees on a hot day with a space temperature of 72 degrees. As I
work with automation systems quite a bit, it is not unusual for the discharge
air to rise 6 degrees between the discharge of the fan unit and the discharge of
a VAV box. And this is with a duct insulated with 2" of fiber. This is not a
negligible heat gain, especially when working with large air handlers.
Paul Diglio
Nick,
The website consists of my notes for teaching at IIT and case studies based on my experience at SOM. It was a hasty part-time effort. The website is paid for until December 18.
Some of the eQUEST models will generate 1000s of pages of Warning & Caution messages. I have never used eQUEST on a real project. The DOE21E models (that I might add later) are based on actual projects and might be more useful. See http://bepan.info/design 5 - Energy-Savings-Options
The Excel programs were created to solve specific project problems and they are not generic programs (could be developed into generic) that could be used on other projects.
I will continue adding the stuff lying around in my computer (from all the junk created when I was working at SOM and teaching at IIT) to the website until December 18. I still have to add my notes on Electrical, Lighting, Plumbing, & Fire-Protection.
Those with an engineering degree and a few years AE design experience do not need the information in the website. It is intended for teaching architectural students.
Varkie