5-ton monster of a heat pump
Hi, Sorry for the book ahead of time.
Ive recently wrapped up an extensive remodel on a 2800 sq ft residential in the greater Hartford area of CT. I had my mechanicals done by a relative that replaced the existing furnace and AC unit with a high-efficiency propane unit and a grossly oversized heat pump. The Furnace and heat exchanger coil were installed late fall of 2015, while the heat pump wasn’t ordered until the following spring. When I discovered that he was planning to install a 5-ton unit, I insisted on downsizing the unit over fears of short cycling and higher humidity levels. Unfortunately, as the coil inside was already installed, and sized for a 5-ton outdoor unit. So the install went ahead.
Having run the unit for 2 years now, I have found the unit does not short cycle during the summer months. This has me baffled due to its size and the cooling loads experienced. On days with outdoor temperatures in the 80-90s, when we use the a/c, I typically see runtimes of 15-30 minutes, with relative interior humidity maintained in the 40’s, sometimes upper 30’s. Recovery times, when we come back after being gone all day, are typically an hour or more. I maintain an ambient temperature of 75, which sets back to 78 when the house is unoccupied.
Regarding the house itself. It is a contemporary, with mostly cathedral ceilings, and an open living concept. The entrance is in the walkout basement, and all 3 levels are connected via open stairwells, while the living room is completely open to the loft. Walls are 2×6 advanced framing with your typical batt insulation. Basement is finished with r-13 batts and steel studs( eventually replacing this with rigid foam.) I have improved areas of the house that had the exterior walls opened during the renovations, using dense-packed cellulose. The main cathedral ceiling has an estimated r-50 between batt insulation and deck mounted foil faced-polyiso. I don’t have exact blower door figures, but they were slightly better than your typical 1970’s constructed house. Not stellar, with room for improvement, but nothing that would significantly increase my cooling and heating loads. All the windows have been replaced with double pane casement windows, where it made sense and fixed pane windows everywhere else. I had a u factor of .27 to .24. Frames are pultruded fiberglass. Overall, the house is not an energy hog, fairly mediocre from an energy conservation point of view.
My question is; with a unit so oversized for the structure, how is it not short cycling? I’m unsure if its a design issue with the ductwork, or the volume of air to condition( due to cathedral ceilings.) I wanted to get the communities thoughts on what might be the culprit. Although I’ve been happy with the performance of the system, bills haven’t been high, I’m curious why it performs the way it does.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
David,
First of all, we can't determine whether the unit is right-sized or oversized until someone performs a Manual J load calculation. For more information on load calculations, see these articles:
Saving Energy With Manual J and Manual D
Calculating Cooling Loads
Who Can Perform My Load Calculations?
Second, some 5-ton heat pumps are variable-speed units. So when the unit operates for 15 minutes and then shuts off -- which you think is normal -- it may be operating at low speed.
Measure the duty cycle, not just the run times, on days when it will be near Hartford's 1% outside design temp (=88F). That will give you a much better idea of the overall oversizing factor. (This weekend's and next week's forecast seems like it would be a good time to take those measurements.)
A 2x oversizing factor isn't necessarily going to create a short cycling condition, since the house has thermal mass, and the thermostat has some high/low hysteresis built in. But it can lead to lower latent load management. So far this season the latent loads have been low to negative, but that low latent condition will change over the next 4-5 days.
https://www.wunderground.com/forecast/us/ct/hartford?cm_ven=localwx_10day
Martin,
I can get a manual J done, but regardless, due to the interior relative humidity levels I see when the unit is running, I have my doubts that it is short cycling. Also, when trying to recover from a setback event, it takes quite a bit of time to bring the temperature down 3 degrees. I know a properly calculated manual J will help me see a better picture of the cooling loads on the house.
Dana,
I can get duty cycle data going back for the past 2 years. I am using an ecobee 3 thermostat, so I have records of exact run times and calls for cooling. Would a snap shot of that be helpful?
The outdoor dew points in southern New England have been unusually low this spring/summer, but you're right- it's not short cycling. If your minimum run times are in the 10+ minute range (look at your historical data, make a histogram of the run times), it's not short cycling. But that doesn't necessarily mean it isn't oversized by a large factor.
The duty cycle snapshots that would matter are only for periods when it's within a few degrees of 88F (your 1% outside design temp.) You'd have to look up hourly weather data to correlate it with for the historical ecobee data to be meaningful. If it's running 15 minutes on/15 minutes off when it's in that range it's 2x oversized. If it's 15 minutes on/30 minutes off it's 3x oversized.
But even without hourly weather data if you look at the ecobee history and can't find ANY four to six hour periods where the duty cycle exceeded 50% it's probably more than 2x oversized.
A 2800' house could easily have cooling load of 3 tons, but it could also be less than 2 tons. Normally it would need to have very good shade from mid-day on for a 5 ton air conditioner to be 3x oversized or more for a 2800' house. Typical cooling loads for a decently insulated house that size would be about a ton per 1200-1600', which would be about 2 tons, give or take. But there are many exceptions. See Allison Bailes' square feet per ton plotted against house size on dozens of his company's Manual-Js.
https://www.energyvanguard.com/sites/default/files/styles/panopoly_image_original/public/square-feet-per-ton-air-conditioner-sizing.png?itok=vsJxOobH
You'll see there were three in those sample in that size range that were a bit under a ton per 1000', which would put it at ~3 tons for a 2800', but there were also a couple at about a ton per 1900', which would make it ~1.5 tons.
Your likely load is somewhere between those extremes, which means you're probably looking at ~2x oversizing, not ~3x.
For this coming week's weather you should be able to nail down the sizing pretty closely using Allison Bailes' method, using hourly data from a nearby Wunderground weather station, and your ecobee data:
https://www.greenbuildingadvisor.com/blogs/dept/building-science/how-tell-if-your-air-conditioner-oversized
Thank you for the additional comments, I will keep an eye on things and report back with what I find in a few weeks.
If you're getting RH% down into the 30's at times, there's a good chance that the airflow is much lower than a typical 400cfm/ton design value (for mixed/humid climates). This would have the effect of reducing sensible cooling capacity...perhaps significantly (10-20%?).
I suspect this is a fairly typical condition for North American heat pumps and AC's....oversized on paper, but with low real life airflow = close-to-right-sized sensible cooling capacity, decent moisture control and higher than necessary energy use.
It's also worth noting that a typical "5-ton" (nominal) heat pump or AC usually has no more than 4.25 to 4.5 tons of actually cooling capacity at AHRI rated conditions.
John: You may be reading too much into low indoor RH reported. Outdoor dew points have even occasionally hit the (uncharacteristic) low 40s and high 30s recently in southern New England. At 75F a 40F dew point = 28% RH. On those low dew point days even with very modest amounts of cooling the indoor RH have been hitting the 30s (but not the past couple of days.)
With the ridiculously oversized low duty cycle central air at my house (also a 5 tonner, for about a 1-ton load due to very favorable shading factors) the indoor RH usually hit's the 40s by the end of a cycle. But my loads are so low it hasn't even turned on during the recent very dry-air days, with the kid up stairs running the half ton window shaker while playing video games. There were quite a few such days in the past month, but the coming week that won't be the case. Beginning Friday/Saturday we're looking at a week of 90s for the high (the 1% outside design temp at my house is 83F), with dew points in the high 60s / low70s (feels like the Carolinas or something!)
https://www.wunderground.com/forecast/us/ma/worcester?cm_ven=localwx_10day
I've found my ecobee's humidity sensor is off almost 10% (say 50% vs 40%) from another sensor in the house...I'll need to buy a third one to figure out which one is correct.
In the preferences you can create an offset if needed.
I have based my comments regarding humidity levels on my experience over the course of the past 2 and 1/2 years. John does make a valid point regarding the lower airflow. I can not recall ever hearing/seeing the blower come up to high speed. During the winter the blower will modulate to a much higher speed when there is a call for the 2nd stage to turn on. This has never happened during A/C use.
Regarding humidity levels the past couple of months, if anything, I have had higher indoor humidity than in the past, generally low 50's. This is due to not having to run the unit as often and occupant behavior( I have 3 children under the age of 5 and a wife that doesn't believe in using bathroom vents.)