|
Evaporator sizing - is bigger better?
Now I have my defrost circuit monitoring the temperature on both sides of my evaporator I got to thinking, would a bigger evaporator be more efficient?
Heat pumps obtain their energy from low grade heat (the atmosphere) and 'convert' it into high grade heat using a compressor. The warmer the temperature of the gas entering the compressor the less hard the compressor has to work to compress the gas and raise its temperature. Whilst monitoring my evaporator I have seen variations of temperature difference of between 1C and 6C depending on the ambient temperature. The exit temperature is invariably less than ambient temperature (it approaches it at very low temperatures). So, my thought is, with a larger evaporator there would be extra surface area (and time) for the refrigerant to absorb heat from the atmosphere, hence improving the efficiency of the heat pump. Has anyone tried this? The argument seems sound but does it work in practice? I have a machine with a dead compressor but a perfectly serviceable fan and evaporator and I'm tempted to try it out but would love to hear any feedback on my proposed hack before getting out the cutters and brazing torch. Acuario |
Yes, a larger evaporator will make a heat pump more efficient. With more surface area for heat transfer, and the same compressor being used, the compressor will move more heat due to the lower compression ratio. How much more depends on your mass flow and lift.
The bugger here is matching the flows between the two units running off of one compressor. How easily that can be accomplished really depends on how the individual evaporators were designed. If they both have integrated distributors and txv metering, the job becomes much simpler. But if they don't employ the same metering method, or use cap tubes or electronic valves, they will hunt against each other. |
Acuario,
There is a point where making the evaporator too large and you loose a lot of efficiency. Take this to the extreme to make the point. Imagine a standard 36000 BTU (3 ton) air conditioner with a standard evaporator. Now, put in place an evaporator the size of a house. The gas that fills the evaporator has expanded almost to room pressure and it has long given up all the latent BTUs. Yes, it is cooled slightly more than say an evaporator the size of a kitchen and only slightly more than an evaporator the size of a bed. The point being that the pressure - BTU - cooling curves are not linear, but are S shaped. Evaporators work on the linear part of the sigmoid curve. Beyond that and the change in pressure vs temp curve goes rather flat. Also, imagine the volume of refrigerant you would need!! Does this make sense? Steve |
Yes, makes sense - house size evaporators are hereby scrapped :-)
What I was thinking of though was to link the output side of one evaporator to the input side of the other so connecting them in series - I didn't intend to use 2 x metering devices or running 2 evaporators in parallel, just more evaporator if you see what I'm thinking. I would need to add charge to the system, both machines are about the same rating (around 5.5Kw) but that's ok. So guys, worth a try ? Acuario |
Why not run them in parallel? Wouldn't that give you the same surface area, but less flow restriction (not sure if that is an issue or not)?
|
I think Daox is onto something (evaporators in parallel). About 1/2 the resistance, a lot more surface area - and a bit more refrigerant.
Steve |
Quote:
As far as I can tell, most of the manufactured units are not optimized for maximum efficiency, but rather are optimized for maximum profit to the manufacturer, so there are certainly efficiency gains to be made by an experimenter. I think that your idea is worth trying, but why you would shy away from running them in parallel, with duplicate metering devices? I see this done all the time in single evaporator units... ...so how would the principle be any different with two separate evaporators? If you are seeking more heat from your evaporator, is there any chance that utilizing heat from the ground will work for you? If your soil is easily drillable, there is a lot of heat to be had there. Also, if you could direct sun-warmed air through your evaporator, it could be even more valuable to you than a bigger evaporator. Best, -AC |
I didn't intend running them in parallel as they are different manufacturers, different caps etc. so as Jeff says it may cause a problem with hunting with the mismatched metering devices - I don't have txv's to put in.
Ground heat is out where I live - dig down 10cm and you hit solid rock. The evaporator is already positioned in a sun trap so I'm getting about as much boost as I'm going to get without moving the machine to one of the south facing walls - and I already have a machine mounted there used for heating the dhw. The idea I had forming in my head was to bolt the units together, one on top of the other then feed the output at the top of one evap to the input at the bottom of the second. The top of this then feeds back down to the compressor. A fairly simple plumbing job. I can wire the fans in parallel so they both run when the unit is on. Acuario |
Quote:
-AC |
Both Daox and Stevehull make good points here concerning flow. Parallel flow is the way you want to go with the evaporator side. Assuming the same outdoor temperature at the two units, the flow will tend to be equally divided between the two units when plumbed in parallel. With this divided flow, the refrigerant will spend more time in each evaporator at a lower speed, and heat transfer will increase in both units simultaneously. With similar expansion devices in each unit, balancing the heat transfer is a simple matter. This is how the manufacturers optimize the heat transfer in multi-circuit exchangers like the one in AC's post. Defrost happens faster, but requires a sensor on each unit.
With series-connected units, the flow is constant in both units, and with the added surface area, this flow will increase. The increased flow causes pressure drops which add up to less effective drop in each unit, causing less delta-t in each at a higher flow rate. The heat transfer is still greater than with only one unit, but the units will tend to find equilibrium where one will transfer a lot more heat than the other. Usually, the first unit after the expansion valve will transfer most of the heat, and the other will be left with whatever is left. As temps approach freezing, the first unit will frost up and the load will be transferred by the second unit until it frosts up. Defrost cycles will be enormous. |
In the parallel arrangement, the units can still be balanced against each other to equalize refrigerant flow. They will tend to frost up simultaneously; as the more efficient evap frosts up, it builds up back pressure due to its expansion valve closing.This increases the flow to the clear evap, which will then frost up until it develops back pressure... Process repeats until both are clogged enough to call for defrost. Upon cycle reverse, both fill up with warm gas and the colder one "sucks" more gas until both are free of ice.
Also with series-connected units, there is no real way to balance heat transfer. Due to the single path through both units, superheat control may not be feasible. One unit will always beat the other out of its share of the load. In a water exchanger with counterflow, this effect can be minimized, but in a crossflow air exchanger with equal intake temps, it's nearly impossible. |
Quote:
How is this different from doubling the length of a pipe with a fluid flowing in it? Doubling the length of the pipe will double the surface area, but doubling the length also doubles the friction, so to maintain flow rate you will need higher compressor power. Did I miss seeing that you included "increasing compressor power"? -AC |
AC,
In the series-connected circuit, I agree that there is more friction present. With a cap-tube metered evaporator, the flow will actually decrease due to this added frction and pressure drop, and the cap tube will need to be modified to adequately feed the larger evaporator. If not modified, the increase in suction superheat and compression ratio could literally burn out the compressor. A TXV or EXV metered system will attempt to do this automatically if the sensor is moved to the end of the second evaporator. But with the same size compressor, the added heat transfer area will bring the heat gained in the evaporator to the point where the added pressure drop will literally be "sucked out" by the compressor (by increased mass flow and resulting expansion at a higher leaving temperature). |
I'm impressed with the knowledge and reasoning - and it all seemed so simple - I'm glad I asked the question now.
I'm going to do a bit of research on the two machines to see if it feasible to put them in parallel rather than series - maybe I'll have to invest in a couple more txv's and junk the capillaries as I'm certain they will be mismatched. Acuario |
Have done both series and parallel.
In parallel used separate TXV (cheap on ebay or surplus), totally different evaporators, one conventional air, the other water tube in tube. If cold enough to freeze up the air coil, the water coil handled the entire load (air txv basically shuts off). Like others have said, you do need to use 2 TXV with sense bulb a foot or more ahead of the junction on the suction line. Have also put a 4T air evap on a 2T compressor, which is basically what you are doing in adding a series evap. TXV is so much more efficient than cap tube that you should use TXV for series connected evaps also, but only one evap needed. Another advantage of the series connection is that the coil will not freeze up until air temp is about 35F if evap overall large enough.. Other part is to oversize the condenser, that way part of the condenser works as a supercooler. Only disadvantage there is that output air (or water) temp is lower; but the lower temp means higher efficiendy. |
Something to beware of is oil return. R290 (either as a refrigerant or an additive) easily solves the oil return issue, but connecting the evaporators in series makes it unlikely to be an issue in the first place.
I have used an ejector approach where I divide up the evaporator into two sections. One part goes where an evaporator would go on a normal system. Between the first evaporator and the TXV, there's an assembly known as an ejector (a kind of pump that has no moving parts apart from the fluid itself) which basically creates a reduced pressure zone for the second evaporator. I also have a phase separator between the first evaporator and ejector tube to supply liquid to the second evaporator. I used a way oversized TXV (5 ton in a 1/2 ton system, can be found cheap on the surplus market) to control flow to the second evaporator. (The reason why the second TXV must be way oversized is because the pressure difference across it is much smaller than in normal use. While a R22 TXV normally sees on the order of 100-200 PSI across it, the TXV for the second evaporator would probably only see 10-20 PSI or so.) BTW, the ejector approach was first used in the Prius A/C system. By cooling the air in two stages, dehumidification is improved. I highly doubt it would offer much (if any) advantage for a heat pump, and it would certainly be difficult to design such that it would work both ways. Just stick to a series connection. |
Quote:
Very interesting approach. So, are you saying that you ran air-to-refrigerant AND water-to-refrigerant concurrently in parallel? Did you notice any "hunting" as has been suggested is the behavior of non-identical evaporators in a previous post? Also, are you saying the air-to-refrigerant side shut down when it frosted up? Was the water-to-refrigerant being drawn on the whole time? Reason I'm interested is that I'm cooking up a hybrid ASHP/GSHP and I plan to run the ASHP most of the time but when the temp takes a nosedive, I'll bring up the GSHP side. From the description you gave, it might be possible to allow the system to make the changes... to be self-regulating, depending on frost conditions. -AC |
by the question:
.. you ran air-to-refrigerant AND water-to-refrigerant concurrently in parallel? Yes Did you notice any "hunting" as has been suggested is the behavior of non-identical evaporators in a previous post? Previous post mentioned 'hunting' for a system without 2 separate TXV, I had 2 separate TXV Also, are you saying the air-to-refrigerant side shut down when it frosted up? Yes, since that evap could boil off less refrigerant, the txv on that evap closed down Was the water-to-refrigerant being drawn on the whole time? This was a system I built over 30 years ago, but recall I had a tsat that turned on the water when air tem was 50F or below. Now, having said that, the system I have now operates with 2 separate compressors, which provides redundancy in case of one system failure. The compressor failed when I was out of town, and DW had a hard time keeping warm, so figured to have 2 separate systems. What I have now is I replaced the 4t compressor in my air-air HP with a 2T compressor. COP increased from about 3 to 4.5 or so at 40F. Built the 5T GSHP described in you epic thread as a stand alone system. Above 40F the 2T air-air operates and never freezes with the oversize evap as the coil temp only goes down to 32.1F. The 2T air-air is able to heat the 5300 sq ft house OK at 40F outside. At 40F outdoor temp, and outdoor thermostat switches over to the GSHP by way of a few custom relay and FET circuits. The air-air HP blower operates for both systems, but at different speeds. It is all also intertied with the fireplace, which has water wall pipes and 2 old car air conditioner condensers in the ductwork and a hot water circulating pump. When there is a hot enough fire, the HPs are locked out. May have time later in the week to copy the control schematic and post in a separate thread. |
Larger evaporator
Quote:
Hi; The laws of physics and thermodynamics have a basic rule regarding heat transfer or movement. From hot to cold and directly related to the difference of temperatures from one medium to another and inversely to the insulation value of the medium. Aluminum and copper transfer heat rapidly while wood fiber glass of course insulate' Larger coils on either end of the cycle means more efficiency. Note that most late model ac units outside units are huge. I don't know enough to tell you how to balance the internal workings of this switch. I notice that my unit has 4 banks of coils with a txv to each while old units had basically 1 metering device to the coil. Owen |
Maybe you could build a radiant wall on the exterior of the house. Something similar to those homemade hot water solar collectors. Run a heavy glycol solution through there, have that run through a big HX to act as a giant evaporator. Maybe also combine design of thermosyphon heater to encourage passive air flow across the radiant panels. Make them modular so you can add on more over time.
|
While I can't answer about multiple evaporators, on the conventional split systems I install I usually go a half ton larger on the evap coil. The added surface area slows airflow and you get better heat and dehumidification.
|
Quote:
I could not find any evidence that one ton (two sizes) up on the evap coil results in further improvement. Perhaps 1/2 ton is the edge of diminishing returns. |
Quote:
> one point on the SEER scale." It this relative to a 1/2 Ton heat pump, or a 4.5 Ton Heat pump, or larger, or smaller? Size matters. -AC |
Quote:
What we are talking about here is the outdoor unit heat exchanger being oversized. With the correct refrigerant metering scheme, heat transfer will be increased. This will lead to increased mass flow through the compressor, and higher discharge temperature and pressure. Indoors, this leads to increased condenser temperature and pressure in the air handler, and increased supply air temperature at the registers. Perceived "warmth" effect will be increased regardless of whether you change the indoor coil or not. |
If the indoor coil has more surface area, you might get more EER/SEER out of it but when cooling that extra surface area can diminish some of its dehumidification capability because the coil needs to be saturated before the water starts to run off of it and once you've met the temperature demand of the cycle and it shuts off, all of the water remaining in the coil will evaporate back into the air after the cycle and be return to the building. Taking a look at the AHRI ratings for 1.5 ton and 2 ton units of the Bryant variety in their 16 SEER models, I'm usually only seeing a boost of about 1/2 SEER by jumping from a 2 ton to a 3 ton coil. I'd rather sacrifice a partial SEER point and have better dehumidification and get a condenser and blower that are as efficient as they can be if I'm going with a central air unit because reducing latent heat(removing moisture) is very important, it isn't just about removing sensible heat(reducing temperature).
I currently have a 2 ton and if I went with a whole house heat pump I'd be willing to keep the slight oversize because I can manage to run a single cycle and still be plenty comfortable but a 1.5 ton would cause the temperature to be more constant while its running rather than have it get to 75 to 78 before powering on and have it shut off at 70-72 when it is done for the day. I've found that this approach removes enough moisture in the air to make the next day's high 70's feel very comfortable. I'd only go to a 2 ton for if I wanted a heat pump because I'd get closer to the capacity I'd need for heating without too much cooling oversize for my house, also the HSPF and EER/SEER usually is higher in the 2 ton units for the brand I'm looking at because people who have low loads to use a 1.5 ton aren't usually going to use too much energy in comparison to someone who needs a 3 ton, for example, and it's harder to sell premium efficiency units in that size. |
Quote:
|
Quote:
Granted system design is extremely important and I think a slightly undersized system(rounding down from manual J instead of rounding to the next .5 ton size) is the key to the best efficiency but contractors aren't willing to do it and homeowners generally don't understand the benefit of better moisture control and efficiency of longer cycles. Looking at the expanded performance charts with systems that use TXVs and ECM motors, it seems to me that you can get decent latent removal with the most efficient sensible removal at 375-400 CFM per ton. 350 CFM per ton is usually used for the beginning 30 minutes of a cycle by many manufacturers to saturate the coil and they go to 400 CFM for most efficient sensible removal after that but if you run a long enough cycle, it seems to me that the math works better for the 375-400 CFM per ton range. ECM motors can move the 400CFM per ton without much increase in blower power, the increased load to the condenser might increase power to it but it would be a good use of that power. |
Quote:
|
This might be helpful, some AHRI data that I pulled when doing this research. Bryant equipment.
For the 180BNA024 condenser EER is .4 higher, SEER .6 higher with the same furnace and a 3 ton coil versus 2 ton. For the 116BNA024 condenser EER is .5 higher SEER .7 higher. With the 986 furnace(with a variable speed ECM versus the constant torque ECM of the 925) the gain is .2 EER and .3 SEER for the 180BNA024 condensor. The 116BNA024 condensor the gain is .2 EER and .5 SEER. This is why the efficiency gains seem insignificant to me versus the extra moisture load that the larger evaporator holds over and releases back into the air and the lag before it actually removes it from the building. If you need to use energy again in the next cycle to remove the extra moisture added back with the larger coil you aren't saving energy while its doing that. EER, SEER, BTU, furnace model(40k BTU size), condenser model name, condensor model, coil size. 13.5 19.3 26400 925 Evolution 20 180BNA024****A 37k coil 12.5 15 24000 925 Preferred 16 126BNA024****A** 31k coil 12.5 15.2 24000 925 Legacy RNC 16 116BNA024****A* 36k coil 13.2 18.5 25000 986 Evolution 20 180BNA024****A 30k coil 12.7 15.5 24000 986 116BNA024****A* 36k coil 13.1 18.7 25400 925 Evolution 20 180BNA024****A 24k coil 12 15.3 24800 925 Preferred 17 127ANA024****A* 24k coil 12 14.5 24000 925 Legacy RNC 16 116BNA024****A 24k coil 13 18.2 24600 986 Evolution 20 180BNA024****A 24k coil 12.5 15 23800 986 116BNA024****A* 24k coil 12.5 15 23800 986 Preferred 16 CA16NA024****A 24k coil 11.2 13.5 23200 925 13 SEER ENTRY CA13NA024**** 24k coil 11.2 13.5 23200 925 Legacy RNC 13 113AN(A,W)024-D 24k coil |
Sensible to latent heat ratio (humidity removal) closely follows the Saturated Suction Tempature of the evaporator coil. Larger coils and more airflow increase SST, trading better efficiency for less humidity removal. Smaller coils/lower airflow lowers SST, give up efficiency for better humidity removal. Go above 50 degrees and humidity removal will be minimal, go below 35 and you will freeze up the coil. Most systems are balanced well between 40f-45f SST. A TXV will keep superheat constant so most of the coil surface remains active. With a fixed orfice the amount of of the coil that is active wil vary depending on the load.
|
Quote:
If your logic is correct, you have just defeated heat pump heating, and all other low temperature heating strategies. Because heat from an oil fired heating duct can be uncomfortably warm, but does that mean that a heat pump heated house with the same inside air temperature would feel less warm, therefore less satisfying than oil heat? If you wanted to actually enjoy the efficiency gain of a larger condenser, you would need to increase the volume of air through it, to get those BTUs into the room. Or, if radiant floor was your heating method, you would need to increase floor efficiency through lower R-value materials above the heated part of the floor structure, and/or increased pipe density and/or increased water flow rates. -AC -AC |
Quote:
Register temp 100-110f is warm enough to reduce the "cold draft effect" but cool enough for efficient operation. This is where variable speed blowers help, no "cold blast" when the blower first kicks on. |
Quote:
Now, at the registers you have to find the proper balance. Too fast and you get the cold draft feel but too slow and you don't heat the room properly. Around 600 FPM is idea, give or take 50-100fpm. This gives you enough mass to get the heat into the room but it's slow enough that in the occupied zone you won't feel a draft. This is a function of the duct sizing and also the way it hooks to the register will affect how the air is distributed. The ideal system is one that you can't tell if it's on or off. Now with oil or gas heat you have plenty of temperature rise over the heat exchanger 40-60°f in most cases vs as low as 15°f with a heat pump. 60 ° return air + 60° temperature rise = 120° out of the heat exchanger that will be 110°+ out of the vents which will be warm enough that it will feel like warm air hitting you. Air velocities don't matter and most people will like a higher velocity that they will feel hitting them. I haven't designed a ton of systems but I've done them both ways and in the end the house is comfortable. Coming from a forced air gas heater I'm used to feeling the blast of warm air coming out and it takes some getting used to when I'm in one of my rentals with a heat pump. The house is warm but you don't feel the airflow. I believe there is some benefit to a slightly larger evap coil but I don't see any benefit to grossly over sizing them. I also believe the best benefit is for the air conditioning side in humid environments. The units I install that have fixed orifices typically have sizing for a 1/2 ton larger A-coil and also include the orifice for it so you get the proper superheat out of them. They typically do not have sizing listed for anything past 1/2 larger - i.e. they have the direct match and a 1/2 ton larger A-coil. I typically install 2-3.5 ton systems and with conventional split systems the range is 1.5 tons to 5 tons, I can't speak for mini-split sizing. |
AC and SERVICETECH,
Yee haw! You're both right. Because heat from an oil fired heating duct can be uncomfortably warm, but does that mean that a heat pump heated house with the same inside air temperature would feel less warm, therefore less satisfying than oil heat? Most people who have had oil or gas blast furnaces in the past tend to expect that "shot" of hot air from the heater when it first starts up. This is one of the big reasons heat pumps took a long time to catch on, especially in Northern latitudes. As with the whole short cycling fallacy, they perceive a long cycle time of less "hot" air as inefficient and costing more in the end. We know this idea is backwards as a left hand thread, but you will never convince Joe Dirt. More indoor coil surface area is great for heat pumps in heating mode. Lower coil bypass factor will mean lower head pressure with the same discharge air tempatures. Airflow does not need to be increased to make it happen, although higher airflow rates would lead to further reductions in head pressure at the expense of lower temp at the register. Register temp 100-110f is warm enough to reduce the "cold draft effect" but cool enough for efficient operation. The most important thing here is delaying the blower startup until the indoor coil has time to preheat. If the blower is not delayed, and starts up with the outdoor unit, the heat coming out of the registers will take time to develop. This goes against the previously mentioned fallacy. Joe Dirt will complain about it. Oversizing the indoor coil will prolong this lag, more so if the airflow is increased. Even though the unit may gain 10-15 percent in COP, the perceived lack of "hotter" air at the registers may become a point of contention with the homeowner. The unit may cost much less to run every month, but other than that, the unit seems not to perform any better than before. |
Many homeowners tend to flip the switch over to emergency heat when they feel temps below 90f coming from the registers or notice the unit running almost constantly. Variable speed blowers solve a lot of the low discharge temp issues since they can ramp up slowly and keep the outlet temp at optimum level. If they didn't charge such a price premium for them... You could also install a time delay/coil thermostat/head pressure switch on a conventional air handler to help with "cold blast".
|
First $1,000 electric bill will cure them of flipping the emergency heat on... :)
I haven't installed a HP system without a delay and variable speed blower. Combined with proper duct sizing I haven't had any problems. The only "problem" I ever have is from people with fossil fuel heaters that expect a blast of hot air when the heater comes on. I had one nit-wit I had to deal with on that who claimed the heater was broken yet the house was almost 80f when I got there with outside being in the 30's. "But I don't feel hot air coming out... I keep upping the thermostat but I don't feel hot air coming out... Why am I sweating?" Wasn't too bright that one... |
I get a few of those also, think something is wrong when 120+ air isn't coming from the vents.
The biggest surprise to me is how few contractors install heat pumps in our area. 75% of all-electric homes use straight resistance heat, 25% heat pumps. Many claim "heat pumps don't work in Oklahoma". Then there are other contractors who install Dual Fuel setups to get rebate money, even though with current gas rates they make no sense. |
Theoretical Heat Pump efficiency vs Delta-T
3 Attachment(s)
Quote:
The formula for max theoretical heat pump efficiency looks like this: ...where all temps are in Kelvin. So I did a little spreadsheet with a typical date set. I'm more interested in GSHP than ASHP and forced air heating (which in my humble opinion is obsolete), so the data set is typical for GSHP scenarios, but the basic principle still applies. Then I graphed Max COP (values are %) vs T warm F (AKA: condenser temp) and here is what it looks like: The first thing to understand is that the efficiency is theoretical, and will never be reached, but it is the theoretical efficiency that lies behind the real world efficiency. Next to observe is that the efficiency gain from decreasing condenser temp from 120 to 110 is not as "smidgen-ish" as one might guess. Even more important is that further reductions of condenser temperatures enjoy an accelerating efficiency curve. So how dd we take advantage of the accellerating efficiency curve? Reduce heating requirements:
When that is done, lower temperature heating becomes possible.
Then you are moving out of the smidgen area and moving into a more exciting area of increasing effficiency. Best, -AC |
| All times are GMT -5. The time now is 08:41 PM. |
Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.
Ad Management by RedTyger