Are Oversized Air Conditioning Systems Really Inefficient? Debunking the Myths

[Cliff_J]

Ok, I keep hearing layperson explanations of the disadvantages of an oversized air conditioning system for a household, say a 5 ton unit is installed in a home that would typically only have a 3 ton. Basically that the oversized system will not exchange the air properly or remove the same amount of humidity that the small system would do if ran for a longer period of time.

Rather than my best guesses as to why this is incorrect, is there any truth to this?

 

[russ_watters]

Yes, it is true - though the specifics are, as always, a little more complicated than just saying if you put a 5 ton unit where a 3 ton unit will be, you'll have problems... The reason is http://ianrpubs.unl.edu/generalag/g626.htm . Let me provide a real-world example:

Last summer, a law firm came to my company with a humidity problem: papers curling on desks, copiers jamming - humidity was about 75%. After a study, we found that the basic reason is that the law office has large offices, conference rooms that are often empty, ect. Ie: a highly variable, but generally light load.

The a/c system was constant volume, with two-stages of cooling. (Translation: while they splurged for a widescreen tv and a water-wall in the lobby, they spent more on those than on an engineer to design them a decent system). That means that if the load was light, only one stage of cooling would kick on and the air sent to the offices would be cooled to 65F or so. Now, when an air conditioner cools the air, it cools the air without removing any (much) humidity until the air reaches 100% humidity - then the only way to reduce the temperature is to pull out moisture. So on a humid New Jersey summer day, supplying 65F air means 65F air at 100% humidity. Warm that air up to 75F and the humidity becomes about 70%.

Take a look at the psych chart I linked: On the bottom is dry-bulb temperature. The temp a regular thermometer reads. On the left is wet-bulb temperature. That's literally the temperature measured when you put a wet sock over the bulb of the thermometer. Its the temperature at which water will condense out of the air. From above - at 65F dry and wet bulb (the very left-hand curve, at the intersection of the dry and wet 65F lines), trace horizontally to the right until you hit the vertical line for 75F. That puts you near the 70% humidity curve.

What that means is that with the summer sun coming through the windows, the 65F supply air is warmed up to a room temperature of 75F, resulting in a relative humidity of 70%.

When two stages of cooling are active, the air gets cooled (and de-humidified) down to 55/55. Tracing right from the intersection of the 55/55 lines until you get to 75F gives -- a comfortable 50% humidity.

There are 2 basic solutions: First, and most preferable, simply reduce the quantity of supply-air when you don't need as much cooling. When you do that, the air flows slower through the coil and is cooled more: you get roughly the same amount of cooling with half as much air at a lower temperature. Essentially you changed the way the cooling is being utilized - providing less volume, but more dehumidification.

The second solution is reheat. If you cool the air down to 55F, then immediately reheat it to 75F, you get all the dehumidification without making the space cold. The drawback is obvious: you waste energy through the use of heating coils. However, there is often a way to do this via some sort of heat recovery.

Now, for residential systems, its not as simple as just saying oversizing will result in too much humidity because most are single stage. They do, however, always have multiple fan speeds (sometimes easy to change, sometimes not). If humidity is an issue but capacity is not, slowing down the fan will lower the supply-air temp and improve humidity. They also don't run all the time. So even if they do get the air down to 55F, if they are only on for 5 min out of an hour on a cool but humid spring day, there is only so much dehumidification they can do. Generally, residences have highly variable humidity even with a properly sized system. That said, its guaranteed to be worse if not properly sized.

In offices, air conditioning is used almost year-round. In your home, only in the summer. In the spring, especially, humidity can be a real problem because if its only 75 during the day, people don't run their air conditioning. If its not on, it can't dehumidify. There are some good residential systems out there that can deal with these issues better (Carrier makes a good one), like commercial systems do, but very few people spend the extra $1500 or so for them.

 

[quark]

No, that is wrong. The main problem with higher capacity AC is lower temperature and you may have to provide reheat other wise chill out. Power consumption will also be high.

A typical psychrometric process, unless sensible cooling, includes cooling down the air to saturation depending upon the apparatus dew point and then removing the moisture. This is at the coil exit and when this air flows into the controlled space, heat and moisture is added. The slope of the line from the coil exit condition to the required room condition is calculated from the sensible heat ratio(SHR) and the process is designed to suit this slope. If the cooled air exactly follows this path you need not require any reheating. But sometimes this line never meets the saturation line and in that case you have to cool the air to some intermediate DP and then should go for reheating.

What you say is right only if you are not condensing any moisture near cooling coil and removing only sensible heat. A higher capacity AC will cool down the air to much lower temperature and as the difference between DBT and WBT decreases there is more RH in the area. Sensible cooling and reheating to maintain RH is not a good practice over a period of year where you have humid conditions during monsoon. It is always better to cool the air to saturation(generally based on ADP) and carefully design your system to match the SHR line.

This link gives you excellent description about various psychrometric processes

http://www.coolit.co.za/psychart/

Regards,

 

[Cliff_J]

Thanks a ton Russ! (yes, a little pun there too ) Very informative, and explains the complex nature of achieving a comfortable environment. I did not know things like how the dew point was different from the wet bulb temperature at anything other than saturation. So now I know you can have cool air that has even greater relative humidity than the warm air if the A/C coil is unable to operate at a low enough temperature to reach the dew point. But now my mind is off wondering how much of an efficiency penalty there would be to have a system that ran the coils at something like 50F on purpose to remove an larger amount of humidity and heat at the same time from the circulated air like in your first solution.

Here's an odd twist in my home. I switched from the expensive disposable 1" thick 10 micron filters to a electrostatic filter that's 6' thick and noticed a huge difference in airflow. I didn't measure it before/after with an anemometer or anything, but feeling with my hand its substantial difference that's more obvious than the fan/heat/AC speed settings for the blower motor. And the system's ability to keep the house cool seems to be working much better this year compared to last year (thus far). I'm noticing a much lower electrical bill and I hadn't expected anything other than better filtration so its not a compliant but a curiousity. And according to my $10 temp/RH meter its roughly the same humidity level of 45%. Hmmm, always nice to pile unscientifically collected data of questionable accuracy on top of more of the same and try to understand anything...LOL

Would I also be correct to assume from psychrometrics that above some 'comfortable' temperature that we would use the wet bulb as the human measure of how subjectively warm/hot air is since we use evaporative cooling? As in the cliche expressions of 'dry heat' for Phoenix versus something like how people refer to Atlanta as Hot-lanta?

 

[russ_watters]

No, that is wrong...

What you say is right only if you are not condensing any moisture near cooling coil and removing only sensible heat. A higher capacity AC will cool down the air to much lower temperature and as the difference between DBT and WBT decreases there is more RH in the area.

Two things: first, it is more complicated than just a yes or no.

A higher capacity coil will not cool to a lower temperature - all are designed for roughly the same output temperature, depending on conditions. They virtually always provide between 52F and 56F by design unless [second thing] they are multi-stage and then the first stage provides roughly half as much as the total.

I didn't want to go too far into the psychrometrics - I already said a lot - but when cooling the air, virtually all of the initial cooling is sensible - ie, you move horizontally from left to right on the psych chart. So if you have only the first stage of cooling, you do no dehumidification: all sensible (what a regular thermometer measures), no latent (de-humidification).

Sorry, got to go to work...I'll post more later.

 

[quark]

when cooling the air, virtually all of the initial cooling is sensible - ie, you move horizontally from left to right on the psych chart. So if you have only the first stage of cooling, you do no dehumidification: all sensible (what a regular thermometer measures), no latent (de-humidification).

I think you meant cooling from right to left

I do agree that the final temperature depends upon coil set temperature but there are two culprits that generally bother us when we go with higher sizes. One is capacity control and the other is system response. At any point of time, the higher capacity coils tend to absorb heat from the control areas at a greater rate than generated and this causes uncomfort.

The case history you mentioned shows that with extra capacity coil the RH comes down. Further, when you don't require to dehumidify the conditioned air, you need not go upto the point of saturation. Just go left to the required temperature but this will give you high RH.

We should better discuss this issue with some numbers.

Filter issue

You didn't specify whether you are getting less air flowrate or more air flowrate when you replaced your filter. If the system pressure drop is low, fans operate to the right on their performance curve and thus you get more airflow, but power consumption by the fan increases. When the pressure drop is high, the fan operating point goes to the left and you will get low flowrates and so low power consumption.

 

[Cliff_J]

I noticed a large increase in airflow. The sound of the fan in the airflow does not seemed to have changed much (yes, unscientific data again) but the system also seems to run less. Coupled with an extra 18 inches of insulation in the attic and maybe some window tinting I think I'll have a pretty energy efficient home in comparison to what it was when I moved in. Hmm, maybe someone in the chem eng forum could tell me the formulas for a homebrew low-E coating for the glass...

I think the thing that led me the most astray on how the AC system works (other than looking at it as a linear model) is that one time back in my apartment the evaporator coil froze up into a solid block of ice. From that I had mistakenly thought the coils operated at a lower temperature. But after looking at a psychometric chart and factoring in how the coils operate in the 50s, it makes far more sense how dependent the system performance is on the other variables inside the envrionment. With the performance of the outside consdenser coil so dependent on the outside air temp too, it now seems pretty amazing how well the average residential system works given that most are installed on little more than an educated guess and have basically only an on/off state of operation.