Heat Pump Defrost Cycle Design 1

[w2jo]

My job is to interface various HVAC and other equipment to Building Automation Systems. I have just (about) completed the interface of a Carrier Infinity Heat Pump to a standard Building Automation Controller. This involved complete replacement of the existing Infinity HP controller so as to allow detailed control and monitoring of the HP by the building automation system.

Every thing is finished and working fine, but I am thinking that the DEFROST CYCLE for the HP is more complex than it needs to be. Maybe someone can educate me.

What I implemented (in accord with Carrier's Manual) is:
1) Set minimum defrost interval to 30 minutes at power up.
2) If minimum defrost interval timer has expired AND OCT (outside expansion coil exit temp) is less than 32F for 5 minutes, then initiate defrost cycle.
3) If the prior defrost cycle took less than 3 minutes (for OCT to reach 65F which signifies "defrost complete", set minimum defrost interval to 90 minutes. If it took 3>6 minutes set next interval to 60 minutes. If more than 6 minutes set next interval to 30 minutes. (This 30/60/90 selection is the continuous running time UNTIL the next defrost is ALLOWED to occur. The defrost cycle is not run unless #2 above is satisfied.

I have noticed that the typical OCT coil exit temp to Outside Air temp when running in HIGH HEAT MODE is about 8 degrees F after stabilizing for 5 minutes. I also have noticed that the OCT temp difference to outside air temp increases as the outside coil ices up as one would expect. When the coil to Outside Air temp is about 12F, the coil is getting "fuzzy" with ice.

It seems to me that if I keep a long term trend of the delta-T (temp difference, Outside Coil to Outside Air in HEAT MODE) and use that as a baseline, I should enter the defrost cycle about when the delta-T gets to 12F or so.

Is there anything I am missing with the above protocol?

Also, Is there a good rule of thumb criteria for a temperature below which a heat pump should not be run? Some have told me 40F, others 37F and others 32F. It seems to me that as long as the compressor AMPS are within specs and the defrost cycle is working the machine should be capable of operating independent of any set minimum temperature.

My criteria for terminating heat pump operation might be if the defrost cycle took more than (say) 20% of the "between defrost cycle" HP running time then stop using the heat pump until the temperature rises. Anyone have any thoughts here?

Many thanks for your helpul comment.

Joe

 

[gepman]

w2jo
I am going to preface this by stating that I have no experience with heat pumps except my little through the wall Carrier unit which cools and heats my barn which I converted to a home office. However heat pumps are basically air conditioners so I will give you my input for what it is worth since no one else with more experience has replied.

Based upon my defrost experience you can use any reliable coil frosting criteria that you can find to initiate defrost. This is difficult for some applications (like -45 deg. F frozen food processing) but easier with yours. I am not sure if you mean that OCT is the outside coil leaving air temperature or the outside coil refrigerant temperature as it leaves the coil with superheat. I suspect the latter since as the coil ices there would be less load and the suction pressure would decrease lowering the temperature.

The minimum temperature for heat pump operation depends on the compressor, the refrigerant, and the discharge conditions. You won't have a problem with amps since as the suction pressure decreases you have less mass flow of refrigerant due to the decreased density of the refrigerant (check typical compressor curves or tables to see the power trend). You will probably have a problem with compressor overheating since the compression ratio will be increasing and the mass flow of refrigerant (for compressor cooling) will be decreasing. You need to monitor compressor head temperatures. You also do not want to operate beyond maximum compression ratios (usually due to the overheating already mentioned but sometimes due to compressor max pressure rating) and there is no reason to operate once the COP at your suction and discharge conditions is less than one (since electric heat will be more efficient then). Your unit probably has a Carrier or Carlyle compressor so it will be a little more difficult to get operating information on it (unless they OEM it out to other manufacturers). If it was a Copeland, Danfoss, Bitzer, or other OEM unit then you could get full information on its operating characteristics.

There should be some application engineers for some of the compressor manufacturers that can give better comments. I am just an ignorant end user of the equipment.

 

[w2jo]

THANKS for your post.. The information about the compressor characteristics was new. The compressor in this particular unit is a Copeland so I can likely get data on it.

In the absence of more definitive information, (and remembering one of my professor's comments to the effect that a well designed and executed experiment is worth 10,000 expert opinions) I have the following experiment operating on this unit.

I have instrumented:
1) Outside Air Temp at the Heat Pump.
2) OCT (temp at the discharge end of the Outside heat exchanger.
3) SUCT (temp of the suction line (larger copper tube) as it enters the outside enclosure.)
4) LIQT (temp of the liquid line (smaller copper tube) as it enters the outside ecnlosure.
5) AMPS (compressor running amps.
6) FANA (Heat Pump Fan Amps.)
7) DIFF (differential air pressure between the outside and inside of the heat pump enclosure box.) I am thinking that this (cheap and easy to obtain) parameter may well be a key measurement in determining when the coil is freezing up.

I would like to have access to compressor suction pressure, but since this parameter is not easily available in the general case, I am not instrumenting that for now.

Now I am awaiting a "big freeze" here in the Atlanta area. I am logging each point once per minute into the controller memory and I will watch the coil itself and the parameters and over time, I think I will be able to establish a criteria for some defrost cycle criteria that can be of general use with this sort of equipment. My goal is to design an algorithm for the controller that will automatically adapt to "most any" rooftop or even residential heat pump so as to (intelligently and with efficiency) allow the unit to operate to the lowest efficient temperature before switching to auxiliary heat.

I have also implemented a COP calculation for the heat pump and have included an entry for the "cost per therm" for the auxiliary heat so as not to accidently run the heat pump below its cost effective temperature even if it is able to do so.

If anyone has any suggestions, comments, helpful hints, please DO. I am an EE stumbling around in ME territory here and can use all of the help I can get!

Thanks!
Joe

 

[gepman]

w2jo
if you can't get the information on the compressor then someone can run it for you through the Copeland selection software. I would measure the head temperature if you plan on running at lower ambients than recommended. Copeland has a recommended location to measure the temperature if you research their website (if you have an OEM password).

 

[imok2]

w2jo, for Emerson/Copeland go here:

http://www.emersonclimate.com/Divisions/CC_001164.htm

 

[tbedford]

...

the temperature below which the heat pump should not run is called the balance point. At this point the heat pump dose not produce heat economically cheaper than your supplemental heat source. A compressor runs with approx. 30 moving parts (UNLESS A SCROLL).
When you have your compressor capacity curve and your structure heat loss, the balance point is a straight line to the point at which you don't save money anymore...hence activate the supplemental heat.

t

 

[gepman]

w2jo
send me the Copeland model number, refrigerant, and the condensing temperature (which will be the air leaving coil temperature plus the coil degree of approach (or temperature difference). I will let you know which is lower, the point at which the COP=1 or where the compressor either overheats or exceeds the lower limits on suction pressure that Copeland has put on its operation. I will post the compressor performance table so you can see it.

 

[w2jo]

Thanks Gepman!
I will get the model number for you. I do have the "balance point" curves for the compressor but frankly, I did not know how to put it to use. Would a copy of the graph be useful to you?

I have installed a 0-1" differental pressure sensor across the inside to outside of the HP outdoor unit box. At 100% fan and no icing the pressure drop is about 0.11". The max icing I have so far developed increased the pressure drop to about .12" and had no measurable effect on the (OAIR - OCT) reading which is nominally 7F in high heat mode. Just now, the low temps here are in the range of 30F so I am not getting much data. The outlet temp differential at 48000btu (high) setting is staying about 20 to 22F so that looks fine as well. I have run it down to OAIR temp of 30F and performance was as above. Compressor current draw was about 11 amps and fan current about 1.8A in the heat pump unit at 30F.

I think the differential air pressure across the coil will prove to be a good indicator of icing.

Thanks for the help.. I will get back with the compressor model.

 

[EmeraldCoastHVACR]

The important thing to remember is the HP's outdoor coil operates 20 to 25 degrees below ambient to absorb heat.

If time & temperature controlled, the HP will initiate defrost with a coil temp of 25 degrees(45-50 deg ambient) by closing the defrost relay contacts.

Increasingly, manufacturers are using an air pressure switch to measure the pressure drop across the coil resulting from ice accumulation on the coil by wiring the air pressure switch with timer & temperature sensor.

Defrost termination normally occurs at 50 degrees at the location of the temperature sensor -- defrost normally should last for 10 minutes.

 

[EmeraldCoastHVACR]

temp of outdoor fan is not required since the outdoor fan should not be running in the defrost cycle.

In the defrost mode, the HP's 4 way valve reverses to cooling mode making the outdoor coil the condensor -- outdoor fan must be off(and auxillary heat running to prevent cold air entering the conditioned space) to insure efficiency.

 

[EmeraldCoastHVACR]

tbedford,

The balance point isn't neccesarily the point that the HP should not run(unless the auxillary heat is gas or oil) -- the balance point is temperature of the structure were the HP needs auxillary heat. If the auxillary heat is electric, the electric can be initiated in stages while the HP is absorbing heat from the outside air.

Typical air-to-air HP's have COP's of 1.5 at an ambient of 0 degrees -- still better than electrics 1.0

 

[w2jo]

I surely do appreciate your input since my last posting!! I have been doing quite a bit of experimentation in the past week. For anyone interested, a draft of my experimentation is posted at

http://gpsinformation.org/joe/HeatPump.html

I have to wait for some cold(er) weather before I can fully quantify my recommended defrost algorithm.

 

[w2jo]

With regard to the comment by EmeraldCoastHVACR, I must agree. This particular heat pump is cost effective (and fully rated) down to -3F. Provided of course you can keep the outside evaporator defrosted! It is indeed "cost effective" all the way to -3F.

For the Carrier 25HNA948A30 (a nominal 48,000btu/h rated heat pump) that I am using, the output is rated at 16,000btu/h at -3F with total system KW input of 2.3. Thus, the per therm cost of this heat at -3F is just $1.15. This still beats today's NG, Oil and Electric Heat costs. BUT.. Obviously you are going to have to have supplemental heat on a day like that!

 

[w2jo]

With regard to temp drop from one end of the evaporator coil to the other: I am measuring about 6 to 7 degrees F on this Carrier 25HNA9 48,000 btu coil in the steady state in both HIGH and LOW heating modes. It does go up to about 15F differential when the unit first starts, but within about 5 minutes, the differential is down to the 6 or 7 degrees (with an ice free evaporator).

 

[gepman]

w2jo
You still haven't given me the model number of the Copeland compressor.

If you are really interested in heat pumps my company has done a lot of research for electric utilities on cold climate heat pumps. Unfortunately the results are for use by the utilities since they paid for the work. However here is a link to one of the first in commercial construction. If I get permission to release our white paper I will.

http://www.gotohallowell.com/assets/DS_White_PaperLR.pdf

As you can see they designed it how a good refrigeration engineer would do it when faced with a very high compression ratio, they went to a two stage compound system.

 

[w2jo]

Hello Gepman,
I was not able to get a cross to a Copeland number. Instead, the Carrier Rep sent me a document with all the curves and tables I needed for the 25HNA9048 (48Kbtu/h) unit so I am all set. I do appreciate your help and offer of more help. I still have a lot to learn I am SURE!

I read the Hollowell article. Most interesting! I knew that we were not very deep into recovering the maximum efficiency from the heat pump's refrigerant cycle. I am delighted that the ME guys are hard at work on the problem for colder climates. I am fortunate that I have both a heat pump and NG backup. Many citizens do not have NG and propane and oil are not far from the cost of Resistance heating.

Thanks again for the White Paper. If you have any more such tutorial info, I would be delighted to read it!
Joe

 

[EmeraldCoastHVACR]

Another point on HP's:

A rule of thumb is usually your heating btu/hr will be around 2 to 2.5 times your cooling btu/hr.

It's kind of interesting that if you buy a HP with variable speed motors & fans and size it for your heating load, you'll save a lot more money over the course of a year than if you sized the HP for cooling(which most contractors do)
and used auxillary heat in the winter.

The 4 ton HP with VS fans & motors will use comparable amperage at similar cooling loads when compared to a regular 2 ton HP, but would have 48,000 btu/hr for heating.

 

[w2jo]

Hmmm.. I notice that for the 25HNA9 48Kbtu/h rated heat pump:

In Cooling Mode, the cooling rating at 350cfm/ton, 72F inside ambient, 96F condenser ambient air, the rating is 51.16Kbtu/h.

Then for the Heating mode, the heating rating at 350cfm/ton, 70F inside ambient, 47F evaporator (outside) ambient is 48Kbtu/h.

I have never looked at heat pumps in this detail before so other heat pumps may well be quite different. The Product bulletin on the Carrier 25HNA9 "Infinity Ultra" can be found at

http://gpsinformation.info/joe/25hna9-2pd.pdf

There is a lot of useful technical information for engineers in this product bulletin.

 

[gepman]

The capacity ratings will all depend on the condensing and evaporating temperatures for each case. I took a look at the Carrier product bulletin. The compressor is almost certainly a Copeland "Ultra" (which for them means it has two capacity points, 67% and 100%). The model number is most likely a ZPS49K4E. I didn't go through the performance table of the Carrier and the Copeland to match them (since Carrier does not give the evaporator and condenser degree of approach [or temperature difference] for you HVAC guys) to try to match it. It could be Copeland's next larger size.

I have attached both a performance table for the compressor and Copeland's suggested operating envelope. It shows what I was trying to say at the beginning of the thread that as the suction pressure drops the allowable condensing pressure drops, most likely due to overheating. Remember the gas cools the motor and the heads. On their Discus compressors you will sometimes see fans on the heads to cool them. Sometimes they will liquid inject the scroll to cool it (which of course reduces the efficiency). Industrially we would either water cool or thermosyphon cool the oil, not that practical for a home unit.

Also Carrier states that this is a "two stage" unit. It has two capacity steps but it is not a "two stage" unit since it does not have two separate compression steps. I had this same discussion with Hallowell (see my post above regarding the cold climate heat pump. Although he has a true two stage system he was calling the high stage compressor with two capacity steps a "two stage" compressor. I finally gave up on the discussion because it probably is easier for the average consumer to think of two capacity steps as two stages instead of two compression steps. It may confuse an engineer though.

EmeraldCoastHVAC is right in that most people will size a heat pump on the cooling load but due to the greater spread between the evaporating and condensing pressures during heating (especially when it is very cold) the compressor will have less capacity. The heating load does not necessarily equal the cooling load. Both loads should be checked. That is why the two step or other modulating type compressors are good in a heat pump since it can modulate to the load which may be different between the cooling and the heating case. You can't use a VFD on a Copeland scroll since it is a compliance type scroll and relies partly on the speed of the orbiting scroll to keep the seal.

 

[w2jo]

Thanks gepman.. You are just FULL of interesting data! I was fascinated by the description of how they achieve 2 "stages" of capacity in this scroll compressor. They have a valve that they can open part way up the scroll to "unload" the compressor. I gather this means they simply do not achieve the full capacity pressure differential and that this unloads the compressor motor and reduces the capacity. My first thought was that the SEER would be worse in the 67% (low) capacity mode. But the SEER numbers (IF I assume that the Carrier capacity data tables are correct) are within less than 1 SEER unit if I go back and forth between low and high speed under a constant Outside Air and Inside Air temp and with constant inside duct pressure. Of course, in this scenario, the inside delivered supply air temperature does show the appropriate change from high to low capacity.

I will read your attachments and see what they are about.. THANKS for sending them along.

Joe