Design day for Appendix G Baseline sizing runs

13 posts / 0 new
Last post

I have been letting eQUEST determine baseline equipment capacities using
the weather files selected by the wizard, rather than creating 99.6%/1%
design days as allowed by G3.1.2.2.1. Does anyone recommend using design
days for LEED/Appendix G baselines? If so, how do you determine
DRYBULB-RANGE, or do you set it to zero to run the same DRYBULB-HIGH
temperature 24/7/365? Does the design day method typically maximize
baseline energy consumption compared to TMY/TMY2 weather files?

Thanks,

William Bishop, PE, BEMP, LEED(r) AP

Bishop, Bill2's picture
Offline
Joined: 2011-09-30
Reputation: 0

Hi Bill,

I don't recommend using design days instead of the weather file at this
time. I have heard in the past that the HDD and CDD schedules that are
automatically generated don't actually work for sizing if you don't live in
California. I can't remember the exact details of that rumor but perhaps
someone else will. I also have heard in the past that it is better to use
the weather files as using the design day method tends to underpredict the
size of the systems. That's all we need! More hours loads not met, yikes!!

So it would be a significant amount of extra work for no real gain.

Cheers,

Carol

cmg750's picture
Offline
Joined: 2010-10-05
Reputation: 0

Hi Bill,

I've created custom design days in past projects based on the 1% CDD and 99.6% HDD and it has actually helped reduce unmet load hours by increasing coil sizes, especially in DOA systems. Chapter 14 of the ASHRAE Fundamentals book will give you the climatic data for your location. You might need the electronic version as I don't believe every location is printed in the hard text these days. The data also shows which month is the hottest and coldest. The data the bottom half of the page gives temperature ranges for drybulb and coincident DB/WB for each month. The difference in ranges each month will depend on your location. Here in Boston, the hottest month (July) has a DB range of 15.2?F and the coldest month (January) has a range of 13.1?F

I've also come across instances where the wizard somehow hard-codes erroneous temperatures causing many unmet load hours and troubleshooting will not get you anywhere if eQuest is using the wrong design temperatures. Now, I make a point to check the CDD and HDD of all of my models.

John T. Forester, P.E., LEED AP

jforester's picture
Offline
Joined: 2010-12-21
Reputation: 0

John:

In reference to the ASHRAE 90.1-2007 HVAC form that needs to be uploaded to
LEED, I interpret the Mandatory Provisions, page 2 entries for 1% Summer DB, 1%
Summer WB and 99.6% Winter Temp to mean that the model must use these design
temperatures.

Do you agree?

Paul Diglio

Paul Diglio's picture
Offline
Joined: 2011-09-30
Reputation: 400

Hi Paul,

Yes, I agree. I'm just not sure how often the HDD and CDD inputs get reviewed/commented on by the USGBC review teams. Overall for most locations, the default weather file method in eQuest will get you very close to the correct values.

John T. Forester, P.E., LEED AP

jforester's picture
Offline
Joined: 2010-12-21
Reputation: 0

Classification: UNCLASSIFIED
Caveats: NONE

I don't know who writes the equest program, but is there any expectation of
getting a Monte Carlo type simulation?

What is Monte Carlo? It is a method of stating the probability of an event
then running simulation. A great example is the occupancy in a room:
example [80% chance of being empty, 10% of having 5 people and 10% chance of
having 9 people. (From 8a.m. till 5p.m.)]

Then each hour it randomly selects either (0,5,9) people based on the
probability.

(The only problem would be each run would have unique results, but over the
full year the deviations would be minimal.)

It would really help with meeting rooms, bathrooms, and all the other random
events that can better be expressed as a probability rather than a set number
at a set time.

Currently when you have 10 offices and occupancy is 10%, there is a 1/10th of
a person in each office. Monte Carlo would say that each office has a 10%
chance of a person being present, and the person would be there or they would
not.

Just a question. We make a living making stuff up, this is just another way
to get closer to making up better stuff.

Classification: UNCLASSIFIED
Caveats: NONE

Jeurek's picture
Offline
Joined: 2010-10-07
Reputation: 0

The only problem would be as you said the unrepeatability. You could always make a complex schedule and put 5 people one day for a few hours then 9 then etc... Probably wouldn't be worth the time though.

Charles Land's picture
Offline
Joined: 2011-09-30
Reputation: 0

Monte Carlo methods are usually more for large quantities of repetitive simulations, still to your point of driving at an answer that carries a higher statistical reality. Energy modeling software is essentially not written that way, but the engines behind the software could be used to handle it. If it is only occupancy you want to adjust with the Monte Carlo analysis, you could use eQuest to create the input file language and then write a script in some programming language to do the Monte Carlo analysis and run the model through the DOE2.2 engine. I would guess that doing that would require the DOE2.2 engine files from the developer (go to http://doe2.com ***NOTE: no "www").

The other option is to use the Monte Carlo analysis combined with some statistical determination model to figure out the best schedule to input into eQuest as the schedule. Chances are that for those particular room types, the difference between the Monte Carlo'd results and your educated guess will be minimal compared to the total energy use of the building. Also, keep in mind that TMY weather files are statistically generated typical year weather based on approximately 30-years of weather records. The accuracy of a forward looking energy model will be constrained to the accuracy of those 30-years being representative of the year the model is projecting.

I think everyone on the list can appreciate the desire for the model to be accurate, but the expectation of accuracy for the energy modeler must be equal to or less than the expectation of accuracy for a meteorologist, right?

Jeremy R. Poling, PE, LEED AP+BDC

Jeremy Poling2's picture
Offline
Joined: 2011-09-30
Reputation: 0

Not a bad idea but. a true monte carlo analysis relies on random number
generation, which would yield different occupancies every time you ran the
simulation and therefore different energy results. Which output would you
submit for code compliance? The HVAC industry has simplified/approximated
this by implementing diversity factors. Not to mention the additional
layers of complexity you'd have to add to the software, modeling standards
etc.

Anthony Hardman, PE

Anthony Hardman, PE
ahardman's picture
Offline
Joined: 2011-09-30
Reputation: 0

You might enjoy this paper that will help you quantify the change in modeled
outputs that result from changes to an input variable.

http://www.hvac.okstate.edu/research/Documents/Spitler_Fisher_Zietlow_89.pdf

Although not quite the same thing that you are after, you can perform this
evaluation with quite a bit fewer iterations and get some idea where to
focus on refining the model.

Use BLAST for extra credit ;)

*David S. Eldridge, Jr.**, P**.**E**.**, LEED AP BD+C, BEMP, HBDP*

*
*

*From:* equest-users-bounces at lists.onebuilding.org [mailto:
equest-users-bounces at lists.onebuilding.org] *On Behalf Of *Anthony Hardman
*Sent:* Wednesday, January 12, 2011 3:55 PM
*To:* 'Eurek, John S NWO'; equest-users at lists.onebuilding.org
*Subject:* Re: [Equest-users] Monte Carlo (UNCLASSIFIED)

Not a bad idea but? a true monte carlo analysis relies on random number
generation, which would yield different occupancies every time you ran the
simulation and therefore different energy results. Which output would you
submit for code compliance? The HVAC industry has simplified/approximated
this by implementing diversity factors. Not to mention the additional
layers of complexity you?d have to add to the software, modeling standards
etc.

*Anthony Hardman, PE*

David S Eldridge's picture
Joined: 2011-09-30
Reputation: 2000

Thanks everybody for your replies!

John, Paul,
Are you referring to this document?:
http://www.ashrae.org/docLib/20080827_9012007_6HVACForms.pdf
While I don't think it's a big deal, I disagree that the use of this document requires the baseline model to be sized using the design day information provided on this form, for three reasons. One, this document is optional for LEED, not required, according to the Beta version of the EAp2 online template that I last used. Two, while I can only guess the intent of the form, the design day section of the form appears to be for reference only (containing other general information such as name, address). I do not see any wording in ASHRAE Standard 90.1-2007 that requires 1%/1%/99.6% temps to be used for mechanical design. Three, this document is for the proposed design. Appendix G specifically says a weather file can be used to size the baseline equipment.

Regards,
Bill Bishop

Bishop, Bill2's picture
Offline
Joined: 2011-09-30
Reputation: 0

Hi Bill,

The simplified compliance path (document referenced below) only becomes an option for projects where the building is two stories or fewer with a gross floor area less than 25,000 sft.

That said, Section G3.1.2.2.1 (page 180 of 90.1-2007) states the following:

Sizing Runs. Weather conditions used in sizing runs to determine baseline equipment capacities may be based either on hourly historical weather files containing typical peak conditions or on design days developed using 99.6% heating design temperatures and 1% dry-bulb and 1% wet-bulb cooling design temperatures.

It seams clear now that either method is acceptable for App. G compliance.

John T. Forester, P.E., LEED AP

jforester's picture
Offline
Joined: 2010-12-21
Reputation: 0

John & Bill
I like using a cooling design day with specific CDD schedules that set scheduling lights, people, equipment to be full on (1.0) throughout the day. If the design intent is to be able to cool a full classroom of students any time of the year, then I make DOE-2 size to that condition, even if the anticipated and model usage is reduced in the summer. That works for the zone air flow rate but I recall having to massage the system and/or plant sizes back down with sizing ratios.

It often feels like an art to avoid penalty and unjustified credit for right sized equipment. Any other suggestions for auto-sizing baselines of projects with very diverse use spaces like classrooms, conference rooms?

Paul Riemer

Paul Riemer's picture
Offline
Joined: 2011-09-30
Reputation: 0