Minisplit comparisons
blamus20
| Posted in General Questions on
Is there a somewhat comprehensive comparison of minisplit systems on the market today? There’re a lot of talk between Mitsubishi and Fujitsu, and a bit of Daikin. I’m particularly interested in the less popular systems like LG, Klimaire, etc.
What are the main differences if they all have models that claim to do 20+ SEER and heat down to below 0 degrees.
My system will be for Colorado climate zone 5a, and I’m hoping to install the units myself and then have a licenced HVAC guy do the refrigeration lines.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
Bernard,
The most comprehensive guide is the product guide for cold-climate air-source heat pumps put together by NEEP. Here is the link:
http://www.neep.org/initiatives/high-efficiency-products/emerging-technologies/ashp/cold-climate-air-source-heat-pump
Click the first link on that page (the Excel spreadsheet).
This listing is not very user-friendly, but it's detailed.
For a more user-friendly overview, see:
http://www.greenbuildingadvisor.com/article/how-to-buy-a-ductless-minisplit
Thanks for the spreadsheet. which columns are the most important - I'm most interested in their efficiency ratings at low temperatures. Is there a column that compares that specific metric?
Bernard,
There are 8 relevant columns. As you probably know, "COP" stands for "coefficient of performance." (Google it to learn more.)
The 8 relevant columns have the following titles:
COP at minimum capacity at 47 degrees F
COP at rated capacity at 47 degrees F
COP at maximum capacity at 47 degrees F
COP at minimum capacity at 17 degrees F
COP at rated capacity at 17 degrees F
COP at maximum capacity at 17 degrees F
COP at minimum capacity at 5 degrees F
COP at maximum capacity at 5 degrees F
You need to care about CAPACITY at your 99% outside design temp than it's efficiency at that temp, and more about the EFFICIENCY than capacity at your mean wintertime temperature. It's only at/below the 99% design temp for 4-5% of all winter hours, whereas it's at or above the mean temperature at least half the time.
There are many heat pumps that will still operate below 0F without damaging themselves and have a specified capacity at some lower temp, but the capacity will be a lot better for models using vapor injection scroll compressors than the typical warm-weather type heat pump compressors, and that affects it's efficiency at average zone 5 average winter temps too.
The SEER isn't really relevant for a heating dominated climate either- the HSPF numbers are more important, and not directly related to SEER. Heat pump designs can be optimized for either cooling or heating around a particular temperature range. There are heat pumps out there with barely legal SEER numbers that still have good to excellent HSPF numbers.
The NEEP spreadsheet also has capacity & efficiency columns at much lower temps for heat pumps still specified at much lower temps than +5F. The low-temp specified is listed in column BB and it's max capacity at that temp is in column BE. There are also specs for the minimum modulation at the min-temp (BC) and the COP efficiency at both min & max (BI & BK, respectively.)
Study the spreadsheet and you'll note that at any temperature, the COP for any given unit is usually higher at it's minimum modulation level than at it's max, but that the difference shrinks at lower outdoor temperatures. When modulating to match the load it will be somewhere between those, and it will vary by the actual load which is highly (but not solely) dependent upon temperatures.
When selecting a mini-split for heating in a cold climate it's also important to consider the level of local distributor & technical support. A really great & efficient mini-split that manages to crap out in the dead of winter that takes 30-90 days for the repair/replacement parts to arrive from Taiwan or Thailand instead of 1-2 days from the regional distributor can be a bigger problem than the up-front price or nameplate efficiency. All vendors have repair part supply chain issues when the unit is 10 years old, but some will have it from day 1. Popular models from popular vendors will likely still be repairable in reasonable time frames at least a few years beyond the warranty period if there is reasonable local distributor support. But even the most popular versions sold primarily online from offshore distribution are a bit of a gamble, even within the (usually shorter) warranty period.
Thank you for the pointers. My summer design temp is 91, and winter 0F. Though I know I will be heating a lot more than cooling. My whole house Manual J add up to 62k btu/h. I'm hoping to find a 5 (or more) zone capable unit so I can use ductless minisplits. Is there a "can't go wrong" series/models I should be looking at? Im tempted to just go with Fujitsu since they seem to be all the rage for heat pumps. Their Halcyon line up can have up to 8 heads, which is more than enough for me. Question is if I need the XLTH models.
The down side to multi-head solutions on a single compressor is that the heads don't modulate with load. You can set the fan speeds up/down, but it will always cycle on/off, which is less efficient and less comfortable. A mini-split head married to it's own modulating compressor can change both the compressor speed and blower speeds to adjust to the load, and run very long, nearly continuous and quiet cycles at low speed.
The non-modulating aspects of multi-splits also leads to comfort problems if a head is located in a low-load space. Even a half-ton head is usually more than 2x oversized for a typical bedroom, and at 3x+ oversizing even the bypass refrigerant running through the head when it's "off" and other zones are calling for heat/cooth can sometimes overheat/overcool the room, even when the blower isn't running.
A heat load of 62,000 BTU/hr @ 0F implies a fairly large house- 4000 square feet or bigger, or a very leaky, less insulated house? (Or maybe that's the load if you sleep with the windows open? :-) ) How confident are you of those numbers?
Large houses can also run into limitations on refrigerant line lengths when it's all done on a single compressor. Mitsubishi deals with this somewhat with a "branch box" concept, feeding all the heads from hub that can be located centrally to the head locations, and a fat pair of refrigerant lines between the compressor and branch box.
With an outside design temp of 0F it's better to go with the XLTH systems, which have pan heaters to manage defrost meltwater, keeping ice from building up inside the outdoor unit and causing damage. (I'm not sure if pan heaters are retrofittable to all non-XLTH versions.) Mitsubishi H2i systems have similar defrost ice management features designed into the outdoor units.
The 8 zone H2i Mitsubishi is only good for 54,000 BTU/hr @ +5F, but a pair of the 3 tonners (4 zone each) is good for 33K each @ -13F, and pair of them would more than cover a 62K load @ 0F. According to the submittal sheets the minimum output at +47F is 7200 BTU/hr each (14.4K total), so a pair of them won't be cycling the compressors excessively if your 62K load number is true. Your whole house load at +47 would be between 15-20,000 BTU/hr. The NEEP spreadsheet indicates a minimum of 22,000 BTU/hr each @ +47F, but that doesn't seem right.
There are no 8 zone Fujistu XLTH units, and the 8 zone AOU48RLXFZ1 is only good for 41,000 BTU/hr @ +5F, per the NEEP spreadsheet well below your 62K calculated load. The biggest XLTH multi-split is a 3 tonner good for 36K each @ +5F, 22K @ -15F. A pair of those would cover your calculated load. The minimum output at +47F is 6,800 BTU/hr each according to the submittal sheets, (12,000 BTU/hr each per the NEEP spreadsheet).
If the NEEP data at 47F is correct running with a pair of 3 tonners from either vendor would have cycling issues, but worse with the Mistubishi H2i than with the Fujitsu XLTH.
Triple-check your load numbers, both the whole house and zone by zone numbers. You may be better off from both a comfort and efficiency point of view with fewer zones utilizing some mini-duct cassettes to heat/cool clusters of 2-4 nearby rooms to better match the zone load to the cassette capacity.
Dana & Martin,
Very informative replies, as usual.
There seems to be a lot of back-n-forth on how variable the minisplits are so they don't suffer from on-and-off switching and evidently the source of the concern is that reduces efficiency/COP. The NEEP link above also has a Specification doc on that same webpage that says:
"Compressor must be variable capacity"
So it's safe to assume all those tested mini-splits, both single and multi, have variable?
One point on energy efficiency that nobody has mentioned is power to run the heat pan / defroster. It's listed in column AY called "Input Power (W)" I see some units with very little Wattage (72W) and others using factors of magnitude larger (3kW)! Is this just for the defroster / heat pan to turn on? Seems like a large range of power usage ...
Justin