AC, Calculate your Actual SEER

Okay, it took me far too long to put this thought together, but yesterday I realized that I could actually make a pretty good calculation of my AC's actual SEER. Now that hot weather has settled in I've actually run the AC a few times and I've been doing a checkout. It's a 1994 unit and has problems that I will likely address this summer. Nameplate on the compressor is SEER 12 for a 4 ton unit (48,000 Btu/hr.) But I've suspected the actual SEER was around 8 based on power consumption vs. estimated cooling load requirements for the home. Turns out I was optimistic and it is running about 6.6. Welcome back to the 1970's. :hippie: It might actually be worse, because I'm assuming it is putting out 4 tons of cooling, despite known problems with the A-coil that will restrict it.

All you need are two pieces of information to do the calc, but one is a two parter:

1. The rated capacity of your unit, this is the easy part. Typically this is expressed in tons where 1 ton = 12,000 Btu/hr.

2. The power consumption of your unit while the compressor/blower are running. This is the hard part and required two steps.
a. What I did was to read the electric meter for a short time period while the AC was running. The disk on mine turns 100 times per kwh so I timed 10 turns. (I had done this for the AC last year for a longer time period and the numbers agreed within a few percent so I know that this was a normal cycle.) EDIT: EEK! No it doesn't turns out that 100 revs = 0.72 kwh.
b. Then I read the baseline for the house once the unit kicked off (had to wait 90 seconds for the blower to shut down too--yours might differ in its time delay.) This power I subtracted from the combined, leaving me the AC/blower consumption. My baseline usage at the time was only 5% of the total with the AC running.

SEER = Btu (cooling)/watt-hour (electricity)

In my case:
SEER = 4*12,000 / 7260 watt-hour = 6.6
EDIT: Based on the correct calibration the result is:
SEER =4*12,000 / 5230 watt-hour = 9.2

So this has me curious...what sort of "real world" SEER's are you folks seeing by the above criteria? This won't work for some of the more exotic two stage or variable speed compressors because they will rarely be running at full fixed load (making value #1 very wrong.) But for those of us running single fixed speed compressors this should give a good approximation.

 

At present, no. But it really doesn't matter for a fixed unit since this is nearly a "best case" treatment for real world cooling capacity. If you look at the govt. SEER test conditions they are for only 82 F ambient anyway (not the ~90 F of my test), and the blower size is not specified. I could go to the trouble of sticking my new anemometer in the ducts to read velocity and temps while measuring condensate output to do a balance, and outside temps too, but it wouldn't matter that much. I expect some departure from the rated cooling load, typically lower...but not quite so much from the power draw.

Depends on what you mean by variable. If it is one of the types that actually runs at varying speeds (not the air handler, but the compression) then one would need to be far more precise. But in evaluating a fixed type system this should be good enough.

Yes, even though the compressor/blower system might be fixed, the actual capacity will vary greatly depending on specific conditions. And the electrical load should vary for some/many of these as well. If you have a 95 F day (vs. the 82 base) then the condensing pressure will be greater and the compressor will work harder.

And if you have a bastardized system like mine with a large blower and narrow nominal 4-ton A-coil you can expect some efficiency hit there since there is no way the design amount of air is making it through the mismatched A-coil frame. This will show up in several ways including extra heat load added by the inefficient blower, extra power consumption that likely exceeded the blower size of the govt. SEER test, plus reduced transfer efficiency on the coil.

Plus mine lacks the TXV at present (mentioned as an option in the compressor installation book.) I'm about 99.9% certain that the compressor was SEER rated with a TXV and a smaller blower...because that is how I would have done it to maximize test numbers. Unfortunately, builders and subcontractors go with cheap mismatches as the norm. They also undersize the ducts most of the time.

That's my take on it anyway.

 

I don't believe that is correct. http://www1.eere.energy.gov/buildings/appliance_standards/residential/pdfs/cac_tp_101105.pdf The Fed. Register is a bear to read but it goes into great detail about the various air handlers and how to apply them. So it doesn't make any sense to ignore their energy input. One could also remove the external fan (1/8th hp on mine) if you were correct. It would be so easy to game this that I don't see how it would be allowed. It would completely eliminate the benefits of variable speed air handlers.

Maybe I'm missing it though. Have you got a source for explaining that it is just the compressor that is rated? It looks like they rate a system with a compressor, and sell the compressor with the SEER rating...but not necessarily all the components.

Don't see that I need to as I can read the compressor/fan draw more accurately from the utility meter than reading current, dealing with uncertainties in supplied voltage and power factor. I have multiple blower power measurements under my belt so it is easy to subtract that out. If I'm introducing an error it is well less than 5%.

 

Impossible to do for any reasonable amount of money/effort. That is of course the problem with most homes. You can't rip out the typical undersized ducting and start over because it is buried within the structure. Besides, my choke point is the narrow evaporator coil with blinded off space on either side. The coil has to go, the question I'm facing is whether or not the compressor is going to survive.

What the test reveals in my case is that the compressor is pulling about 80% more than it should be to operate. Once the worst of the current heat passes I'll have the AC guy take a closer look at it. He is stretched thin right now (as I expected.)

At this point I suspect that with the existing ductwork trunks I could put in a smaller AC unit, nearly double the efficiency, and still get about the same effective cooling load. Since the main ductwork is limited by existing geometry, it seems best to work backward from there rather than try to shoehorn in too large of a system.

p.s. A-coil is one of those terms I've picked up from the AC folks. It is shaped like an A, so I'm not surprised by the description.

 

It is my understanding that this type of meter measures real power (watts) not apparent power. You could have a large reactive load but the meter wouldn't show it. Now if you have a large enough reactive component and the utility finds out they could charge you for it. They do this with large commercial customers at times. (I remember that as being something the EE professor would go on and on about in that circuits class so long ago. Seemed to be obsessed with it.)

Might be fun to test some time. I've got enough CFL's that I've measured PF's on that I could do a before and after to see if it is reading VAR or real power. The CFL's are typically in the .6 to .65 PF range IIRC.

 

Ah! Thanks for that, I stand corrected. This is great because it fixes some apparent accounting errors that had been bothering the heck out of me. My vampire draws add up nearly perfectly to the base load now! And this cuts my blower down to ~3/4 hp.
In my defense, the "7.2 kh" is on the meter off in a corner, not the disk, but says nothing about what it corresponds too. Of the many different instruments I've had to determine calibrations before I've not come across anything quite so eccentrically calibrated or labeled. Roughly 100 turns looked right for 1 kwh, but I had not done a 100 turn check vs. the dials to verify. That's what I get for not doing a rigorous check, I know better than that.
This raises the SEER up to 9.2, which is bad in the vintage 1994 mediocre way, but not in the "imminent failure" way. Whew, that simplifies things! (My neighbor's much newer AC started making a hideous knocking sound at the start of the season...so I'm a bit antsy about this old unit.)

 

Do you mean they are going to charge for apparent power rather than real power? If they just read VA then they will have apparent power. If they apply PF then they will be charging for real power.

I'm wondering what they would actually charge a residential consumer for reactive power? It would seem that their own gear would recover it since it isn't really consumed. There would be some resistive loss in their lines for the additional current, and some cost for correcting for the reflected power with capacitors. But it is not being consumed within the residence--converted to work/heat.

The following is a good discussion of reactive power for non-EE types (that would be me) PFC decoded

Hey, I would love to be able to step in and recover reactive power that someone else was being billed for... It's like charging someone a fee for making change.

 

What?? I thought it was robot sheep you were working on...

p.s. We really need a sheep smiley...