[Equest-users] Oversizing..AHU or Plant?

[Arunabha Sau]

Thanks Bruce. Great reply.
I completely agree that modelers need to have a good technical knowledge.
Oversizing AHU is justified now. I appreciate your help.


Much thanks,
Aru


On Tue, Jul 26, 2011 at 7:28 AM, Bruce Easterbrook <bruce5 at bellnet.ca>wrote:

>      The intent of 90.1 and using a modelling program like eQuest is to
> size the equipment closer to the actual demand.  The 15% and 25% variance
> factors were used in the past before the computerized modelling we are using
> today.  Doing load calculations buy hand or using a spreadsheet made for
> much more deviation between the calculations and the building.  Less things
> were accounted for as well.  By providing more certainty using a modelling
> program we should be reducing these safety factors.  These factors are still
> a best guess and the designer must keep all this in mind.  The next step in
> the process is specifying the equipment for the building.  So if the
> modeller has a 15% SF on the equipment and because most equipment is built
> with discrete sizes, the specifier adds another 20% by picking the next size
> up you can see how things start to chain together and get out of control.
> Next step is bidding, the specifier uses Trane as base bid but Carrier wins
> the process and submits shop drawings.  They are normally required to
> provide a superior bid, there is another 8%.  Remember as well the second
> and third increases are now compounding, we are not 43% over sized but
> probably more like 50%+.  My point is the original design and SF should be
> used from start to finish.
>     Next question is where to apply these factors.  They should be on the
> air handlers.  If your fan is too small and your coils are too small what
> you get supplied from the plant is irrelevant.  You physically do not have
> the equipment to handle the diversity.  So the method is to account for the
> diversity in the end units, and have the plant at 1.0 or less.  Your SF on
> the units flows through to the plant.  Note I mentioned "or less".  Consider
> a building with a north south split.  An AHU on the north side will only see
> diversity due to weather.  A south side AHU will see diversity due to
> weather and sun.  You may consider different diversity factors for each.
> Throw in 10 buildings of different sizes and orientations you can see the
> effect in the deviation on the plant using old methods.  eQuest due to the
> through nature of the model can account for all this.  Can is a critical
> word, it all depends on the model.  There is still diversity in the model.
> The amount depends on what the money people want to spend modelling.  LEED
> is the attempt to control these factors, to set a minimum quality level for
> the model.  It is now realized this is not enough.  Original LEED buildings
> on the majority do not perform as the model suggested they would.  Now
> commissioning is the new buzz word.  As if that is the only reason.  If your
> units are too small commissioning won't fix the problem either.  If your
> units are too big the system will work but not efficiently.  Part of the
> problem is the process, hence the proposal that modellers be PE's.  That may
> help but it wouldn't solve many of the problems.  The model is important but
> is only effective if used as a reference start to finish.  The safety
> factors come from using eQuest and doing project reviews.  Comparing the
> actual building performance to the model and taking a look at the allowances
> for variance.  There is only one group responsible for this whole process,
> it is the mechanical engineer who has assumed responsibility for the
> building project.
>     The short answer is there is no rule of thumb for the safety factors.
> More experience allows you to guess better but it is still a guess because
> every project is different.  Model feedback is spotty.  The safety factors
> should be applied at the terminal equipment not at the plant.  Mostly, 15%
> for cooling and 25% for heating are too big.  In most cases .5 cfm/SF is too
> small.  But that is where I start.  Get the model built and running, let it
> run on defaults, let it determine the initial sizing, thoroughly review the
> sim output, then start tuning it.
> Bruce Easterbrook P.Eng.
> Abode Engineering
>
> On 25/07/2011 07:47 PM, Arunabha Sau wrote:
>
> Howdy!
>
>
>  So, I just saw a project which has applied its oversizing factors (1.15
> for cooling and 1.25 for heating) to the plant rather than the air handlers.
>  ASHRAE 90.1 Appendix G page 180, G3.1.2.2 Equipment capacities  does not
> specify where the oversizing should occur.  Have you seen this method used
> before?  Is it an acceptable and accurate way of meeting the baseline
> modeling requirements?
>
>  I appreciate your help.
>
>
>
>
>  Much thanks,
> Aru
>
>
> _______________________________________________
> Equest-users mailing listhttp://lists.onebuilding.org/listinfo.cgi/equest-users-onebuilding.org
> To unsubscribe from this mailing list send  a blank message to EQUEST-USERS-UNSUBSCRIBE at ONEBUILDING.ORG
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/equest-users-onebuilding.org/attachments/20110726/a6c012c9/attachment-0002.htm>