ASHRAE PRM - HVAC Loads Low

[Eng-wintersmith]

I’m working on an early phase LEED/ASHRAE PRM model.

The building loads (both baseline and proposed) calculated are too low to be believed. I work at a Mechanical Engineering firm and have 10 years energy modelling experience and 5 year HVAC design experience. The loads as simply lower than any expectations.

My, inputs, and the PRM report indicate that the internal loads inputs seem reasonable:

Lighting = 0.94 w/sf
Receptacle = 0.45 w/sf
People = 275 sf/ppl

But the load reports indicate the autosized design flow is 0.3 cfm/sf - 0.4 cfm/sf and the cooling load is between 13 and 15 btuh/sf (800 - 920 sf/ton).

These HVAC loads are simply too low. No practicing mechanical engineer is ever going to size his units this low. I do not believe it’s appropriate to size the ASHRAE Baseline Units this low. My (and industry) expectation is about 0.75 – 1.0 cfm/sf and ~400 sf/ton.

In fact, since the min turndown flow for the Baseline VAV system is 0.40 cfm/sf, it is effectively a constant volume unit.

What could be causing such low load results?

[farahghanem]

you can check the loads breakdown using vista and can check for internal and external gains to identify why the values seem low.

The software calculates the loads according to your inputs so you need to verify those first, and especially the profiles to ensure the building gains are operating as expected. Also usually the rules of thumb used by mechanical engineers add 20-30% safety factor. I must admit I have run into such problems where there is significant difference between what IES autosizes for the proposed and what the mechanical design is telling me.

[Eng-wintersmith]

Thanks for your response.

What do you usually do when this mismatch is costing performance. I've not typically been successful telling the ME to downsize his system.

In regards to the schedule. In my experience engineering practice does NOT apply schedules to the internal loads during sizing. They would all be applied at 100%.

In fact Trane/Trace 700, and DOE based programs (which would include Blast, Epro, EQuest etc) all support a special schedule day for sizing.IES does not have such a scheduled day. This means the IES calculated peak load could be much smaller than the 'design' load. For example if the envelope peak occurs during a time when occupancy schedule is low.

I see nothing in ASHRAE 90.1 that would require sizing be conducted the way IES is doing it. And the deviation from engineering practice seems to cost me energy performance.

[farahghanem]

IES calculates the peak load, that is including all internal and external gains. So for cooling mode, you would expect the peak to be around july-september...etc

ASHRAE 90.1 does specify that both baseline and proposed must be operating on the same schedules, and also in the user manual they have provided "acceptable" schedules depending on building type. You can use these schedules in absence of any design/actual profiles.

I do not think that IES calculates loads any differently than it should... unless you messed up inputs somewhere.

As for what I do when designers system appears to be oversized in comparison to what IES is telling, I agree you wouldn't be successful in telling someone who's been copy/pasting the same design across dozens of projects to pay more attention to their sizing. However, I would share with them the full set of my inputs and sizing results from IES, and they would either tell me "increase this, and remove that", or "oh we considered almost the same values but x1.2 or 1.3 for safety... at that point I just nod and smile.

It is good to note that the proposed building SHOULD be penalized for wrongly sized design parameters.