Retrofitting Demand Control Volume (CO2 Based) to office Building 1

[yinthenovice]

Okay guys , I've been looking at this option for a 13 story office building in the Midwest .The current system is VAV , with VFD (VSD's on all air handlers) , also use some economizer , but minimum outside air on some floors is set for max load , when floors are currently only 30%- 40% full .All the case studies seem to indicate people save a ton of cash .

Our building control system is state of the art , so interfacing shouldn't be a problem . My only concern is that the CO2 sensors are notorious for drifting , losing calibration that sort of thing in a few months after installation . So some say you may end up not saving as much as expected.

Has anyone had experience with retrofit Demand control Volume ? Any issues , pitfalls before I take the leap ?
Many thanks

 

[willard3]

I have fixed a couple of these by taking them out because the controllers are lousy and nobody in an office building will maintain/replace them when they start to fail.

 

[CptHk]

The owner must understand that to get the savings of DCV, a smaller amount of money will need to be dedicated to keeping the CO2 sensors calibrated/working.

Our college/university clients tend to understand this. I'm not sure with the office buildings.

 

[ChasBean1]

CO2 sensors can tend to drift and are affected by pressure—if used, I’d suggest a minimum quarterly zero and span calibration for them. I’m not much of a DCV fan either, but it can work if you don’t overcomplicate with too many sensors and controls, etc.

The absolute minimum OA flow among AHUs should be just higher than the total, normally operating building exhaust flow.

The way I would do it (this is not the only way, but could keep controls simple): set around 80% of building VAVs to minimum flow. Have AHUs run normally and continue to maintain duct static pressure. Assuming AHUs have return fans, adjust each AHU return fan to track its supply fan by X Hz so that the sum of all AHUs have minimum OA flow set point about 10% higher than the total building exhaust.

If return air CO2 for each AHU is below 850 ppm, simply run the AHUs at this established supply/return Hz offset. If return air CO2 reaches 1,000 ppm, widen the offset of the return fan up to a certain value, say X-10 Hz. So for an AHU set up for return to lag supply by 5 Hz, if CO2 approaches its upper limit, the offset increases up to 15 Hz and scales between 5 Hz and 15 Hz for return CO2 levels between 850 and 1,000 ppm.

The high Hz offset value could be adjusted based on coil limitations or effectiveness in reducing CO2. The OA dampers could be left at a fixed minimum, say 30% open. If there is a separate bank of minimum OA dampers, they could just be left open during all operation.

This would work for most common systems, but it would be different with a) system with no return fans or b) a return system that is flow or pressure controlled at upstream points, in which case DCV or minimum OA control should not be used.

 

[paultg]

Excellent post by ChasBean1; the only thing I would add is to be sure you understand the "total building exhaust". You need to be aware if the building has exhaust system that operate intermittently; such as fans in a cafeteria, if the building has a gym, day care, etc, or if any tenants have their own small kitchens, and how make-up air is provided to those spaces.

Paul G.

 

[ChasBean1]

Good point about the variable exhaust. The most substantial areas like large kitchens/cafeterias should have their own makeup air; not done like this 100% of the time though.

 

[paultg]

The office buildings I have experience on for the most part had direct make-up for larger exhaust souces (hoods); however it was only 75-80% of the exhaust rate; the remaining air was from the house system.

 

[yinthenovice]

hey guys , thanks for the helpful tips .
Willard & Cpthk , your experience is amazing and i'm even moe encouraged because the guys who do maintanance on this building also maintain an attached data center , so are dilligent by nature , maintance will be done on time.

ChasBean , you ARE the man , the only problem is they only exhaust the building using bathroom fans on each floor and these dont have controls on them . Putting controls that vary their flow rate may cause political problems with tenants if they dont feel they are getting enough air if you know what i mean

 

[ChasBean1]

yinthenovice,

First, thanks!

Second, maybe consider exhaust fixed if it's just the TX load as you mention. Don't control exhaust - assume a fixed value and set min OA just higher than that max operating cfm value.

Usually you'd find an absolute minimum OA just exceeding the toilet/general exhaust values in an office building provides an OA value below the 20 cfm per person at max occupancy... a good minimum; a min that can sustain a positive building but one that can vary based on the occupancy as per return CO2 readings.

 

[cry22]

You should install airflow measuring stations on SA, RA and OA. That's the most reliable way to monitor your building OA and pressurization. Because those CO2 sensors will indeed fail you (out of calibration).
The other day,I was commissioning a CO2 sensor and the thing switched the system to economizer mode in a 90F day with a 40 PPM difference in sumulation. Not reliable at all these CO2 sensors.

I am curious to know from maintenance guys out there on how they really perform, any tracked data will be helpful to all here.

Thanks

 

[ChasBean1]

Respectfully disagree. Fan Hz tracking is repeatable and better. OA flow especially can be rocky and hideous with wind.

Just got through with a building with fans having flow indications bounce from 13K cfm to 45K cfm and back and between. These flow stations were at the inlet bellmouths of the plug fans. Flow control could work if there were a time constant and averaging feature, but this application had no such feature. Just bouncy readings resulting in wigged-out controls...

Controls company, one of the better ones around here, can program in these features but will charge a hefty extra fee. Getting things to run in auto is in the original scope; getting things to actually function nicely is WAY beyond scope. They use the commissioning (the process that finds things that result in poor performance) as a spring board for change orders.

 

[cry22]

"They use the commissioning (the process that finds things that result in poor performance) as a spring board for change orders"

Cool line Chass!!

I never had such reading problems with AFMS you mention though.

 

[ChasBean1]

Thanks cry22! Great feedback! They do, and it's a constant battle.

The airflow readings I've seen on different jobs, if not "rocky and hideous" like I said before, for the most part have had "massively questionable calibration," or can be “laden with variable compoundable errors” (as a couple quotes to summarize an overall opinion of airflow tracking control at AHUs).

Drive frequency control has precision of a dart on a bullseye. Airflow control at the AHU level is like those old sixties cars with that major wheel play, where you turn the thing plus or minus 90° just to keep in a g'darn straight line...

 

[paultg]

ChasBean1:
Can you expand a bit on how airflow is controlled via VFD speed? What is purchased to measure air volume (AFMS?) within the equipment and is this calibrated with the respective equipment fan curves?

What about for OA when it is mixed within a mixing box with return air; and there is only a supply and return fan? These are the types of units I come across more often, and we ask for AFMS across the entire intake louvers.

Paul G.

 

[ChasBean1]

Paul G, here's my spin on it:

Example 1) Supply fan controls to maintain duct static pressure and flow is read by a flow station at the fan. The return fan then modulates speed to maintain its flow at the supply flow minus X cfm for a (theoretical) constant outdoor air. On paper, this would be a very good way to keep a consistent min. OA intake. In reality, I haven't seen this work well long term and maintain consistency.

Example 2) Measuring OA flow and using that value to modulate an OA damper or a return fan VFD or to proportion mixed air. Again, nothing against the concept; the issue is use of an unstable variable to try to get stable performance.

What I've seen most often in place are pitot tubes sensing total pressure (ports facing the airflow) and static pressure (ports 90° from the airflow), from which velocity pressure is extracted and flow is calculated. They are usually supposed to be setup and field calibrated and sometimes they are. Long term, they can tend to develop issues with sensing ports clogging, especially in OA and RA airstreams.

To me, these flow stations are useful for orders of magnitude readings or overall functional verification, but not as inputs for controlled devices. For the mixing box application you mentioned, I'd set return fans on a simple Hz offset from supply to ensure a repeatable minimum OA value, then just use the mixing dampers for what they're made for - temperature control.

That’s what I meant from the last couple of posts – hopefully this clarifies my take a little.