estimating static pressure by furnace power draw
Good morning everyone,
I'm pondering how to better inform my filter change intervals on my furnace. I have an 80k Trane XV80 with a 3.5 ton AC, and a 16x25 Honeywell F100 air cleaner. I do not have manometers or the knowledge/skill to use them to measure the ESP, but I do have access to some sophisticated power measuring tools, which I do know how to use properly. It seems that I could measure the power draw of the furnace and use the power & cfm vs. ESP tables in the XV80 specification to estimate ESP. I can pretty easily force the furnace to run in both the low and high stage for heat, as well as the constant fan mode which I understand delivers ~50% of the cooling CFM. No ESP data is given for that few CFM but perhaps I could extrapolate.
Figured I could measure the power in all three modes with the existing filter to estimate the total ESP, then remove it and repeat the tests without a filter to get an estimate of the pressure drop due to the filter. Then repeat again with a clean filter and see what the difference is.
This may be too DIY for the site, if so I apologize. Thought there might be enough of a science-experiment aspect to make it a worthwhile discussion.
Using static pressure to measure the pressure drop across a filter is typically the most common way to determine a filter's cleanliness vs. amp draw.
Understand that as a centrifugal blower with a constant speed (PSC) motor unloads (as in filter becomes more blocked), it draws less amperage, not more. An ECM motor may not behave the same way. The variable aspect would make your quest to measure the filter's dirt level more challenging, as it would even if you were measuring pressure drop across the filter with a manometer and a variable speed motor.
You, the end user, are in the best position to learn over time how often your filter should be changed. I, as someone who has serviced equipment like yours, have learned that two identical homes side by side can have vastly different filter change interval needs. The elderly lady living alone who has maid service and no pets, who only runs her a/c when it gets brutally hot outside, can go months without a filter change, and be no worse off. The family with six kids next door, with two dogs, a gerbil, a hamster, and a cat, who run heating and cooling full tilt year round...once a month is pushing it. Learn what your usage habits are, and pattern your filter change intervals around that.
Building Physics Rule #1: Hot flows to cold.
Building Physics Rule #2: Higher air pressure moves toward lower air pressure
Building Physics Rule #3: Higher moisture concentration moves toward lower moisture concentration.
Maybe. Trane in their Installation Manual does show the power consumption at different static pressures. I think they are one of the only mfg's that publishes this that I've seen. However, their residential IOM"s take a little bit of a google serach just like Carriers. So I suppose you could take a baseline power measurement when the furnace starts up (ID fan running, but blower off) and subtract that.
Wit ha VS motor, as pressure drop acorss the filter increases, the motor will increase RPM to compensate and use more power. In systems with hgih static, a VS ecm motor cna consume more power than a similar HP rating PSC since it's operating range is higher. The fan curve of the PSC has to be designed for the lower static situations where it woudl consume the most pwoer (highest CFM output) since it's speed is fixed. The VS ECM will simply slow down as static drops.
Constant torque is different. With a constant torque motor, the power output to the windings is roughly fixed (strength of the field). The relatinship between static and RPM is inverse. So as static increases, RPM drops and visa versa. So that's better than a PSC, but CFM still varies since fan curves not linear.
Right, Trane shows wattage vs. ESP (from 0.1 to 0.9) for the various air flow settings in heating and cooling mode of the XV80. It's in medium for both low and high stage heat, so I think I should be able to get a decent correlation between measured power draw and ESP.
I'll be using a power logger that measures volts and amps and calculates power factor and wattage every second. Figure I'll hook that up with the furnace powered off, power it on with no call for heat to get the baseline power used by the electronics (probably very low), start it up in low stage, see how that settles, then increase to high stage and see how that settles. Then I can remove the filter, repeat (I'm assuming running without a filter for 5-10 minutes is OK), and then do it again with a fresh filter.
I'll then plot my results vs. the specifications and see how it all looks. It'll be a fun exercise if nothing else.
Out of curiousity, I understand the Bryant Evolution system has some sort of ESP measurement and filter change determination built in. Does it do this with actual pressure readings or is it something like I'm describing here, operating the fan at a known CFM and using the power draw vs. ESP relationship?
dunno, but hope you'll share your findings & process with us.
by any chance, are you an engineer?
"Thought there might be enough of a science-experiment aspect to make it a worthwhile discussion."
you've tweaked my curiosity!
best of luck.
The cure of the part should not be attempted without the cure of the whole. ~Plato
I think on the carrier infinity it measures blower rpm and power. Im very sure there's not a pressure transmitter.
Sent from my DROID RAZR using Tapatalk 2
Minor update. Since I last posted, my furnace started acting up. It was failing to ignite the burners at times, and at others would run for a bit but shut off before meeting the temperature set point. The stuff I could check seemed to be working (the VisionPro IAQ thermostat was working, furnace responded correctly to switching the fan to "On", etc), didn't see anything obvious so I called the guys who installed it. They came out last Friday and checked it and determined that the filter, a 16x25x4 Honeywell MERV 11, was clogged. This didn't occur to me as a possibility as it had only been in place for ~2.5 months. (I'm an engineer but don't pretend to know it all!) We've been keeping the fan in the On position (that and having the vast majority of run time in the first stage has solved the temperature difference problem we used to have across the three floors of our townhouse) so perhaps that's what led to it getting loaded up so quickly.
Anyway, before that happened I had already ordered new MERV 8 and MERV 12 Honeywell replacement filters from Amazon. They weren't in yet, so I grabbed an "FPR 4" 1" Rheem filter from HD to use while I was waiting for them to come in. They got here yesterday.
Here are some amperage readings I took on the furnace's circuit. Used a Fluke 376, didn't have access to the power quality analyzer so I can't determine actual power. My experience is that variable speed motors show a variation in power factor depending on the load, measured amps and actual power are not linearly related. CFM should be approximately 600-620, as these were taken with the fan set to "On" but no call for heat. The furnace is set up for 3.5 tons, 350 cfm/ton, and the Trane docs say the On cfm is ~50% of the cooling cfm.
Furnace on with fan off - 0.2
fan on, no filter - 0.7
fan on, 1" filter - 0.9
fan on, MERV 8 - 0.8
fan on, MERV 12 - 0.9
fan on, old MERV 11 - 2.9
At the least in the future I can spot check the amps in fan On mode and use maybe 2.5 as a filter replacement limit.
I would simply but a testo static pressure tester. Yuo can then test supply,return,filter,evap to find where real problems lay.
hey, a reason to buy a new toy! ahem...tool.
and yes...by the process you described your esitmating process..
it is obvious that engineering is your background.
but I don't mean it in a bad way!
just that our thought process are quite different.
I'm enjoying your posts.
best of luck
The cure of the part should not be attempted without the cure of the whole. ~Plato
I'd like to point out, that based on what I think you are trying to do (a filter change light, basically), you will likely find that a pressure switch across the filter is going to be cheaper/simpler than a current switch. As the filter loads up, pressure drop across it will increase. Set the switch to where you think the filter is dirty enough to change, and you're done.
I finally got around to taking the measurements. I used a recently calibrated AEMC power quality analyzer with a high accuracy 5 amp range current transformer so the wattage data should be very accurate. I can only estimate the CFMs based on the chart in the specifications document and the assumptions that fan "on" mode runs at 50% nominal and that the dehumidification and ComfortR ramps each run at 80% nominal. I used the data for low, normal, and high airflow settings for the 2.5 and 3.5 ton outdoor units. So those are some sources of potential error.
That said I've definitely learned some things and some of the results match very closely to what should be expected.
For the cooling season I have to partially close a damper on the supply trunk feeding the 1st and 2nd floors (this is a 3 story townhouse) in order to keep the temps even and the 3rd floor comfortable. Also, all the registers on the 1st floor are shut most of the way - 1st floor is half underground at one end and walkout at the other and always stays cool. In the winter everything is opened up and the low-stage heat keeps everything even.
For the test I first set the dip switches for 400 CFM/ton @ 3.5 tons (turns out this is more like 390 CFM/ton) and ComfortR active. Then during the 7.5 minute 80% airflow portion of the ComfortR profile I took power data with no filter, a brand new Honeywell MERV8 filter, and a Honeywell MERV12 that had been in use for a couple of months. I repeated these measurements after the ComfortR profile finished and the fan sped up to 100%. I also ran this test in fan "on" mode.
Then I reset the dip switches for 350 CFM/ton nominal (likewise this is more like 331 CFM/ton) and took measurements in the same manner. The furnace draws ~10 watts when the fan isn't running so I subtracted that from the measurements. So now I have power measurements for 6 airflows ranging from 580 to 1,360, for each of the three filter conditions.
The attached figure plots the measured wattage vs. estimated CFM (2nd order polynomial curve fits to be exact). Additionally, I took the airflow and power draw vs. ESP data from the specification document and plotted lines of constant ESP in the same manner. Last, I added some vertical lines calling out the airflows at which the measurements were taken.
So, it appears that with a new filter in place and cooling the system operates at ESP from 0.6 to 0.9+, increasing with airflow. This makes sense as the filter's pressure drop increases with airflow per the figure in the F100 manual. Interesting to note that the indicated ESP difference between the no filter and new MERV 8 is very close to what the F100 manual suggests at the same airflow.
The data has led me to a couple of actions. First is to leave the nominal airflow at 350 CFM/ton. I do not like that a new filter is showing ESP >0.9 when at 400 CFM/ton, I can't imagine that's good for the long run. Second is that I'll probably stick with the MERV 8 filter, though I should do a proper comparison of a brand new MERV 12 before finalizing that decision. Third, I've set the thermostat and dip switches to use the dehumidification on demand feature and no ComfortR, so it'll run at the 80% speed when the humidity is 50%+. It has been working nicely with this configuration.
Lastly, I'll probably investigate ways to bring the ESP down. I'll start another thread on that topic after I get a chance to check a few things out and get some photos of the installation.
something you may not be considering is the load of moisture on the a/c coil influencing the results. cooling coils have a higher static resistance when wet than they do when dry...
The TRUE highest cost system is the system not installed properly...
I am yourmrfixit