AWHP superheat & sub-cooling

[desA]

Originally Posted by Gary
We can use Te,sat=15C and Tc,sat=75C.

desA:
We continue along this track, then.

I can now confirm that if the 50Hz operating conditions are moved up to Tc,sat=75'C, that similar compressor mass-flow conditions exist as for the compressor running at 60Hz, Te,sat=10'C, Tc,sat=65'C. This calculation also allows for efficiency variation (small).

The difference now lies in the pressure ratio moving from 4.56 to 5.99 - with rpm@60Hz = 3500 & rpm@50Hz = 2900. So the compressor runs slower, but pushes to higher pressure.

I will run thermodynamic simulations on these new assumptions to better characterise the Tcomp,disch for the 60Hz & 50Hz conditions.

What would the effect on oil distribution & wear be at the 50Hz revised operating conditions?

Now, at least, the difference in compressor operating window for US & EUR is more clear. Pity it took two technicians to unravel what the best of the so-called best technical experts, for arguably the largest compressor manufacturer in the world, could not work out. Incredible stuff.

(Ironically, I wonder why I'm not really surprised. Saw so much BS in my previous engineering lifetimes, in various scenarios. The suits live well.)

[desA]

An EPR valve does not allow the Te,sat to drop below setpoint pressure. The exact opposite of what we are trying to accomplish.

That's what I had thought. Thanks for putting that one to rest.

I personally like the CPR concept. Wish we could find a more cost-effective producer.

The fan-control trim is starting to look better & better. I'm working on a system which we can use to better test the Tcomp,disch versus fan speed concept.

[mad fridgie]

Gary I have to disagree about a CPR not flooding the evap. Reason. CPR is limiting flow to compressor (at set point) Ct at 70C, liquid pressure sligtly less, TXVs do not close 100%. Liquid will continue to enter the evap, pressure in evap will rise along with the amount of liquid, basically flooding the evap, or should i say a wet evap outlet. so no superheat control.
Charge then becomes a limiting factor

[mad fridgie]

After studing more of this tread, I would suggest that you control your load (fan speed) using a point on the evap (not in the air flow) at temperature close to the limits of the compressor. As this is a slow acting systems a simple PI control loop is required, I would use a discharge temp sensor for saftey control, not for fan control (indirect) For example there is a slight loss of refrigerant, suction superheat will increase, discharge will increase, fan would slow, lowing suction, increasing compression ratio, increasing comp discharge temp and so on.
TXV if you are lucky find one which can give you enough mass flow at lower pressure ratios and can control at high pressure ratios, If you can not, add a second expansion device. TXV to control for high pressure ratios, second to allow additional mass flow. Sol valve can be controlled, simply or with levels of complexity

[desA]

... Liquid will continue to enter the evap, pressure in evap will rise along with the amount of liquid, basically flooding the evap, or should i say a wet evap outlet. so no superheat control.
Charge then becomes a limiting factor

So far, on the lab rig, the TXV has controlled well over the entire heating range, with no sign of loss of superheat.

This has been one fairly consistent factor & may have something to do with the evaporator design size.

The Te,sat drift occurs over the range of the cycle, in line with the thermodynamic imbalance (my perspective), where evap becomes increasingly more oversized, as load climbs (fixed speed fan).

[desA]

After studing more of this tread, I would suggest that you control your load (fan speed) using a point on the evap (not in the air flow) at temperature close to the limits of the compressor. As this is a slow acting systems a simple PI control loop is required, I would use a discharge temp sensor for saftey control, not for fan control (indirect)

The problem with trying to control off a small change variable, is that sensor noise will play an increasing part. The 'D' in PID can make life interesting for this scenrio, I'd have thought.

My thoughts are to use an input signal with a large range for control purposes. The smaller signal can be used to trim, or bias the output.

For example there is a slight loss of refrigerant, suction superheat will increase, discharge will increase, fan would slow, lowing suction, increasing compression ratio, increasing comp discharge temp and so on.

Good thoughts. Thanks for that.

TXV if you are lucky find one which can give you enough mass flow at lower pressure ratios and can control at high pressure ratios,

All good, so far. No electronics in sight, thankfully.

If you can not, add a second expansion device. TXV to control for high pressure ratios, second to allow additional mass flow. Sol valve can be controlled, simply or with levels of complexity

Thanks for that. The secret seems to be in matching the TXV carefully to compressor capacity, over the range.

[mad fridgie]

Sizing the valve is the secret, In NZ we have a much wider ambient -5 to 35C dry bulb.
In your calcs have you allowed for the influcence of varying humidity levels?
We have found humidity is as important in system design
Control loop PI not PID (ambient and water temperature only change slowly?!)

[desA]

Sizing the valve is the secret, In NZ we have a much wider ambient -5 to 35C dry bulb.

In SE Asia, the dry bulb can vary from say -4'C in winter, in the colder parts, to 42-45'C in the warmer regions - some, even hotter, on the odd occasion. In some parts, it can really get very much colder, in winter, however. Talk about temp spread.

In your calcs have you allowed for the influcence of varying humidity levels?
We have found humidity is as important in system design

Yes, very much so. What humidity levels are you typicaly working at over your way? What influences are you observing on system design?

Control loop PI not PID (ambient and water temperature only change slowly?!)

Water & air, no problem. The refrigerant time constants are very much shorter, however.

[mad fridgie]

Being a temperate climate our daily changes are large, trying to design for all conditions is impossible. But primarily we are low ambient so best design evap for low refrigerant pressure drop ensuring highest LMTD, of course can cause problems elswhere, as you have noticed.
Single pass systems do tend to have greater stability (normally a constant Tc) even though i happy to allow for instability on multipass, as averaged COP is greater, and averaged compression ratios are lower

[desA]

Being a temperate climate our daily changes are large, trying to design for all conditions is impossible. But primarily we are low ambient so best design evap for low refrigerant pressure drop ensuring highest LMTD, of course can cause problems elswhere, as you have noticed.

Fair-enough.

Single pass systems do tend to have greater stability (normally a constant Tc) even though i happy to allow for instability on multipass, as averaged COP is greater, and averaged compression ratios are lower

Agreed. May I ask what type of stability issues you see on the multipass systems?

I'd imagine things like TXV corrections, refrigerant waves (refrigerant movement from evap to condenser), would be much the same.

In terms of humidity, ever tried this test point:
Air WB in = 34.12'C
Air DB in = 32.22'C
RH% = 89.4%



How well do your desuperheaters behave?

[Gary]

Gary I have to disagree about a CPR not flooding the evap. Reason. CPR is limiting flow to compressor (at set point) Ct at 70C, liquid pressure sligtly less, TXVs do not close 100%. Liquid will continue to enter the evap, pressure in evap will rise along with the amount of liquid, basically flooding the evap, or should i say a wet evap outlet. so no superheat control.
Charge then becomes a limiting factor

I think you may be confusing the CPR with the EPR. The TXV does not need to close 100%, it needs only to maintain setpoint flow while the heat load increases. In effect, the evaporator becomes oversized. If anything, the superheat may rise a little, in which case a fan control could reduce the heat load and bring it all back into balance.

[desA]

In effect, the evaporator becomes oversized. If anything, the superheat may rise a little, in which case a fan control could reduce the heat load and bring it all back into balance.

Would you consider partially preempting the superheat rise by say reducing the SH value from 7K to say closer to 5K?

I've seen comments of useful SH values for heat-pumps in the range around 5K. The lab machine has a nice vertical 'liquid trap' leg from evap discharge line to join the suction line which runs high, then lowers into the compressor suction.

The fan speed control is beginning to sound like an appropriate trim control strategy, with CPR as the flow peg.

It's nice knowing that you have a fixed point in the cycle to work against. Can push hard until limiter kicks in, then back off on fan speed, as required.

[Gary]

Superheat at 5K is not at all unreasonable and would improve performance. The TXV can control down to about 3K or even a little lower before it starts to hunt and flood.

The downside is that superheat drops with coil temp and could endanger the compressor on an unusually cold day.

[desA]

Superheat at 5K is not at all unreasonable and would improve performance. The TXV can control down to about 3K or even a little lower before it starts to hunt and flood.

The downside is that superheat drops with coil temp and could endanger the compressor on an unusually cold day.

Fair comment.

Ok, we leave SH on 7K, to give a fair operational safety margin.

[Gary]

I fully expect the VIC to increase Te,sat beyond the 19C we have seen so far.

This will bring the CPR into play early on.

If the CPR is in play given our minimum design air temp of 25C, then we have more evap than we can utilize, which tells us we need to downsize the evap... and possibly tells us how much downsizing is needed.

[Gary]

Fair comment.

Ok, we leave SH on 7K, to give a fair operational safety margin.

The better alternative would be to test the system at a colder air in temp than could be expected for the region and adjust the superheat to just above 3K. Then we know for sure what will happen on an unusually cold day.

[desA]

Originally Posted by desA
Fair comment.

Ok, we leave SH on 7K, to give a fair operational safety margin.

Gary:
The better alternative would be to test the system at a colder air in temp than could be expected for the region and adjust the superheat to just above 3K.

Makes good sense - test both environmental limits & set SH accordingly in the range 3K (at lowest ambient), see how it fares. Test again at high ambient, check SH is still acceptable.

It's going to take some doing to cool an intake in my current lab - the walls are full of air-blocks (SE Asian style of building). I'll have to either back a cold outlet onto the heat-pump evap inlet face, or duct back some air from the existing heat-pump to mix in with hot incoming air. Mmmmhhh... I have access to a full environmental test chamber - but this won't be for the current lab machine.

I'll need to figure out a cooling trick for this.

[Gary]

Makes good sense - test both environmental limits & set SH accordingly in the range 3K (at lowest ambient), see how it fares. Test again at high ambient, check SH is still acceptable.

It's going to take some doing to cool an intake in my current lab - the walls are full of air-blocks (SE Asian style of building). I'll have to either back a cold outlet onto the heat-pump evap inlet face, or duct back some air from the existing heat-pump to mix in with hot incoming air. Mmmmhhh... I have access to a full environmental test chamber - but this won't be for the current lab machine.

I'll need to figure out a cooling trick for this.

At this point, leaving the superheat as is allows us to do comparative testing. The final adjustment can wait until everything else is completed.

[Gary]

Somewhere towards the end of our testing we will need to decide between single-pass and multi-pass, based on the COP.

If multi-pass yields higher COP, then switching over is a simple matter of eliminating or bypassing the water regulating valve.

But there is a happy medium available to us, also.

We can determine a reasonable minimum tank temp, say 55C for example. We can adjust the water regulating valve to where the W,out is 55C when heating the feedwater.

If the tank level drops too far, bypass the valve.

When the tank is full, switch to tank temp maintenance mode and bypass the valve.

This way, the water in the tank is never less than 55C unless there is very heavy usage, such that the tank level drops to the lowest (bypass) level.

We could even add a switch on the control panel which allows the customer to choose between highest efficiency (bypassed) or minimum water temp (regulated).

[desA]

At this point, leaving the superheat as is allows us to do comparative testing. The final adjustment can wait until everything else is completed.

Good idea. Ok, we'll hold on 7K, for now.

[desA]

Somewhere towards the end of our testing we will need to decide between single-pass and multi-pass, based on the COP.
If multi-pass yields higher COP, then switching over is a simple matter of eliminating or bypassing the water regulating valve.
But there is a happy medium available to us, also.

We can determine a reasonable minimum tank temp, say 55C for example. We can adjust the water regulating valve to where the W,out is 55C when heating the feedwater.
If the tank level drops too far, bypass the valve.
When the tank is full, switch to tank temp maintenance mode and bypass the valve.
This way, the water in the tank is never less than 55C unless there is very heavy usage, such that the tank level drops to the lowest (bypass) level.

This is very nice logic. To have both options available, with the water regulating valve set to control to the selected value (say 55'C) is a neat idea.

We could even add a switch on the control panel which allows the customer to choose between highest efficiency (bypassed) or minimum water temp (regulated).

Could we take this a step further? Allow that water flow regulating valve to have a set-point also driven from the control panel.
Customer can then select his preferred tank temp & efficiency option, for the system to operate at.

[mad fridgie]

"I think you may be confusing the CPR with the EPR. The TXV does not need to close 100%, it needs only to maintain setpoint flow while the heat load increases. In effect, the evaporator becomes oversized. If anything, the superheat may rise a little, in which case a fan control could reduce the heat load and bring it all back into balance."
__________________
Hi gary any restriction in the suction line between evap and compressor EPR, CPR, Hand valve, Blocked filter, can cause increased liquid in the evap, to what level depends upon the equalibrium state of the system, generally during design these restrictions (pressure drops) are allowed for to allow sufficeient flow to allow other control devices (TXVs) to operate normally, during abnormal circumstances, the CPR, EPR, is the main controlling/limiting mechanism.
Primarily we are protecting against abnormalities and not normal steady state conditions.

[mad fridgie]

DesA, tried 43Cdb, 40Cwb, ran for 4 hours, actually no problems. (test was completed just to see how well the compressor handled it!)
Refrigeration migration is always a problem, liquid reciever helps, but adds cost and increase amount of refrigerant

[Gary]

Could we take this a step further? Allow that water flow regulating valve to have a set-point also driven from the control panel.
Customer can then select his preferred tank temp & efficiency option, for the system to operate at.

We can do anything... for a price.

But we may need to go electronic.

[mad fridgie]

desuperheaters, (not specific to heat pump), very commonly used, have found that as long as the entering water is cool, they are great, if the entering water is hot, then the HX can cause a pressure drop, driving up the discharge pressure and temp, whilst trying to maintain the same Ct, so would think a similar problem may exist on your unit

[mad fridgie]

Just a point of interest, NZ there is aproduct called mahana blue, they use very high St and Ct, they have been given permission from the comp manufactures (maneurope) to use out of normal working envelope, this also may be an option for you look at, speak to your comp manufacturer

[desA]

"I think you may be confusing the CPR with the EPR. The TXV does not need to close 100%, it needs only to maintain setpoint flow while the heat load increases. In effect, the evaporator becomes oversized. If anything, the superheat may rise a little, in which case a fan control could reduce the heat load and bring it all back into balance."
__________________
Hi gary any restriction in the suction line between evap and compressor EPR, CPR, Hand valve, Blocked filter, can cause increased liquid in the evap, to what level depends upon the equalibrium state of the system, generally during design these restrictions (pressure drops) are allowed for to allow sufficeient flow to allow other control devices (TXVs) to operate normally, during abnormal circumstances, the CPR, EPR, is the main controlling/limiting mechanism.
Primarily we are protecting against abnormalities and not normal steady state conditions.

I typically design for ultra-low pressure-drop in the suction line - though sufficient for oil conveyance.

In the case of a CPR addition, I'd not be overly concerned at this stage, especially with its only function being a slight trim-off of entry pressure.

We'll have to see how it plays itself out in the system & may require some feminine massage.

[Gary]

As I see it:

The CPR holds the mass flow at X, because the heat load has risen to X+1. The liquid cannot pool in the evaporator because the heat load is more than enough to boil it off faster than it is flowing in.

This is the exact opposite from the EPR where the mass flow is held at X because the heat load has dropped to X-1. There is not enough heat load to boil off the flow of liquid.

[mad fridgie]

CPR, controls on outlet pressure, If the liquid is at at higher pressure entering the TXV than the CPR outlet pressure, then flow will occur, if the load temp is less than Ct then pooling can occur, maybe I should have included dependent on charge. The liquid will fill the evap to the point where the effective heat transfer is reduced (both liquid pooling and reduce TD).If charge is critical, then Liquid is held in evap, we loose the liquid seal to TXV, massive reduction in mass flow through TXV, head drops, suction mass flow increases boiling of liquid in evap, pressures drop and a way we go again, Systems will reach equalibrium!

[Gary]

As I see it:

The CPR holds the mass flow at X, because the heat load has risen to X+1. The liquid cannot pool in the evaporator because the heat load is more than enough to boil it off faster than it is flowing in.

In this scenario, if the fan speed were reduced, bringing the heat load down to X (or lower), all would go back to normal.

Do you agree with this?

[mad fridgie]

For this scenario, I believe fan speed control is all that he needs, so yes I agree

[Gary]

For this scenario, I believe fan speed control is all that he needs, so yes I agree

How can dropping the heatload below X possibly clear a flooded evaporator? The flow isn't going to increase before or after the CPR. The flow will in fact be less than X if the heat load is less than X.

[mad fridgie]

Do not believe you need a CPR, load control (fan speed), will ensure Te remains within specifications, On start when Te will be high, Ct will be low (presume reasonable anti-cycle protection) minimal additional torque required, so practically not an issue for a short period running out of envelope.

[Gary]

Do not believe you need a CPR, load control (fan speed), will ensure Te remains within specifications, On start when Te will be high, Ct will be low (presume reasonable anti-cycle protection) minimal additional torque required, so practically not an issue for a short period running out of envelope.

I'm thinking either the fan speed control or the CPR could do the job, but in concert the compressor is protected from both overloading and overheating.

Hmmm... now that I think about it, the fan speed control could be configured to protect the compressor in both ways.

[mad fridgie]

quote
Hmmm... now that I think about it, the fan speed control could be configured to protect the compressor in both ways.
Yes Either could be used, I must say that i have never seen a CPR on this type of product?

[mad fridgie]

Non related question, how do you post others quotes?

[Gary]

Just hit the "quote" button on the bottom right of the post you want to quote.

[mad fridgie]

cheers for that

[desA]

Do not believe you need a CPR, load control (fan speed), will ensure Te remains within specifications, On start when Te will be high, Ct will be low (presume reasonable anti-cycle protection) minimal additional torque required, so practically not an issue for a short period running out of envelope.

Let me state from practical experience that, in some cases, fan speed control alone is not sufficient to pull Te,sat down to where the compressor manufacturer would like it to be.

It is very much dependent on the fan type & its achievable turn-down ratio. Depending on the fan speed control mechanism, the fan sometimes just cannot go slow-enough to get the Te,sat down far enough & the system, although dropping much of the way, still hangs a few degrees above the desired value.

Oddly-enough, this can also depend on the day, not necessarily only local inlet temp & obviously is related to a number of variables at that time - local voltage (big issue up here), ambient temp & relative humidity (a biggie, it seems).

This is the reason for going for a belt-&-braces approach - grab & tickle (CPR & fan).

[desA]

...On start when Te will be high, Ct will be low (presume reasonable anti-cycle protection) minimal additional torque required, so practically not an issue for a short period running out of envelope.

Yes, & then some...

[desA]

Just a point of interest, NZ there is aproduct called mahana blue, they use very high St and Ct, they have been given permission from the comp manufactures (maneurope) to use out of normal working envelope, this also may be an option for you look at, speak to your comp manufacturer

Many thanks for the lead. I'll follow up on their product.

I've seen a bunch of compressors go south in hottish climates (rooland) where the designer elected to 'overlook' the finer details. This was very typical of this particular design philosophy. I know the machine well - an updated version of one sits in my lab & is the one we're working on...

[desA]

quote
Hmmm... now that I think about it, the fan speed control could be configured to protect the compressor in both ways.
Yes Either could be used, I must say that i have never seen a CPR on this type of product?

Can you explain what you observe for the LP pressure (Te,sat) over the course of a heating cycle with water heated from say 30'C to 65'C, in say 1.45-2 hours - for a cyclic heatup (pump-around) operation?

This would be for a fixed speed evap fan, with air inlet temp at 32-35'C. Summer operation.

[mad fridgie]

I've seen a bunch of compressors go south in hottish climates (rooland)
I no the machine your ar talking about have a number of thermal overloads strapped to the compressor

[desA]

desA:
I've seen a bunch of compressors go south in hottish climates (rooland)

mad_fridgie:
I no the machine your ar talking about have a number of thermal overloads strapped to the compressor

Very true...

[Gary]

Here is another option, the AEV (automatic expansion valve):

http://www.drillspot.com/products/47...xpansion_Valve

More info:

http://www.parker.com/literature/Aft...Pdf%202007.pdf

[desA]

^ Thanks, Gary.

Now, where would you place that in the system? In place of the TXV?

[desA]

Further to CPR & evap load-up

Are there any instances where TXV bulb can be moved to a position other than directly after the evaporator, on the suction line?

Say if, for instance, a temperature was found to drop off slightly over the cycle, in proportion to the evap internal rise in Te,sat & SH i.e. a related reversing signal e.g. SH' = 7 - a*SH.

[desA]

DesA, tried 43Cdb, 40Cwb, ran for 4 hours, actually no problems. (test was completed just to see how well the compressor handled it!)
Refrigeration migration is always a problem, liquid reciever helps, but adds cost and increase amount of refrigerant

The electronic bits would have surely loved that lot...

[desA]

Here is another option, the AEV (automatic expansion valve):

http://www.drillspot.com/products/47...xpansion_Valve

More info:

http://www.parker.com/literature/Aft...Pdf%202007.pdf

Were you perhaps thinking of using this instead of the CPR?

If so, wouldn't the pressure drop be horrendous?

[Gary]

^ Thanks, Gary.

Now, where would you place that in the system? In place of the TXV?

Yes. Generally these are used where the load is constant and precise evaporator temp is needed. It is probably not something we would use, but it is an alternative to be explored.