Developing a heat-recovery chiller

[rjk_cmh]

I don't really know what to call this thing, but I'm needing some help finding parts that may or may not already exist for this application. It's not just a heat-recovery chiller - I also have a dx coil plumbed in, and the coil can either act as a condenser or as an evaporator by the use of two modulating 3-way valves, and one 6-way switchover valve.

I found the modulating 3-way valve, available from Sporlan. However, I can't seem to find a 6-way valve that is designed for refrigerant use - plenty of them exist for hydronic/chilled beam applications.

The sequence goes like so:

1. Compressor hot gas goes to the first of two modulating 3-way valves.
2. On outlet port 1, we have a brazed-plate heat exchanger acting as a condenser for the hot side of the water chiller system.
3. On outlet port 2, we go into circuit 1 on the 6-way valve.
4. The 6-way valve common ports are connected to the dx heat exchanger, which is remotely piped outside the building.
5. The outlet of 6-way valve circuit 1 goes to the outlet side of the brazed-plate condenser HX, which creates the liquid line. Possibly add a receiver here?
6. On the liquid line, we now have the metering device (EXV).
7. After the EXV, we enter the second modulating 3-way valve.
8. On outlet port 1, we have another brazed-plate heat exchanger acting as an evaporator for the cold side of the water chiller system.
9. On outlet port 2, we go into circuit 2 on the 6-way valve.
10. The outlet of 6-way valve circuit 2 goes to the outlet side of the brazed-plate evaporator HX, which creates the suction line.
11. Suction line goes to accumulator, then back to compressor.

The purpose of this system is to allow for FULL heating capacity while only using PARTIAL cooling capacity and rejecting the rest to atmosphere, or in reverse - FULL cooling capacity while only using PARTIAL heating capacity and rejecting the rest to atmosphere.

The reason I'm choosing to pipe a DX coil outside the structure instead of using a drycooler is because I don't want water lines running outside in the freezing weather, and my water loops can have higher thermal transfer ability by using less glycol or other antifreezes.

An example of one system that is somewhat similar to this is the MultiAqua heat recovery chiller. However, it differs in 2 major ways: it is first of all a package unit with all refrigeration components outside, requiring 4 water pipes to connect, and secondly it cannot throttle the amount of heat rejected or absorbed by the atmosphere DX coil. It only has 3 modes: Heating only (like a standard heat pump), Cooling only (like a standard air-source chiller), or Heat Recovery (outside DX coil not used at all).

My proposed system would have 5 modes: Heating only (like standard heat pump, for when there is no chilled water demand), Cooling only (like a standard air-source chiller, for when there is no hot water demand), Heat Recovery (when there is full demand for both chilled and hot water simultaneously), Heating Main (for when full heating demand is required, but only partial cooling demand), and Cooling Main (for when full cooling demand is required, but only partial heating demand).

In Heating Main and Cooling Main modes, it is not required that there be FULL demand for either heat or cool - only that they be unbalanced demands. In the case where there is partial demand for both heating and cooling, but the demand is less than the total capacity of the system, the compressor could be slowed down with a VFD in order to reduce its capacity.

I still have not figured out the most appropriate way to defrost the DX coil without wasting a lot of energy and reducing the available heating capacity in the winter. It's a necessary evil, but it would also result in the uncontrolled production of chilled water even when there may be no demand for it. I think the solution would be to have a large chilled water buffer tank to absorb the defrost cycle without freezing.

Phew - that was long! I would be happy to hear all criticisms of my grandiose ideas.

[icemeister]

I assume you're envisioning this system for an application akin to to an office building where the core areas may need cooling while the outer spaces need heating. You want to capture the heat generated from chilling the water going to the core and it. Similarly the sunny side of the building will need cooling and the shaded side heating.

When the recovered heat isn't sufficient for heat demand the air source evaporator will pick up some load form outside and supplement the heating.

I suggest staying away from modulating refrigerant side valves and multiport valves. Stick with regular solenoid valve to divert the refrigerant to where you need it to go. The chilled water and hot water should be circulating as primary pumping loops and the building zones can feed secondary zones and modulate the temperatures as needed. Much simpler.

I would be controlling the compressor by looking at both the hot water and chilled water loop supply temperatures, depending upon the dominant mode you're in. Let the water-side controls handle the supply temps to the building. You can reset the primary loop temps setpoints according to OAT or whatever. Again, let the building side handle that...just establish your system limits and be able to function anywhere in between.

The air source unit's defrost is most simply handled by a quick hot gas defrost. The need for defrost can be minimized with a wider fin spacing, but you need some form of heat which must come from with the system unless you use electric or water defrost. A simple electric defrost won't rob heat from the system, but it will take longer, so it still takes otherwise available heat away from the system. An industrial refrigeration type water defrost unit would be fast and efficient, but not cheap. They can use straight city water temps and for the relatively minor defrost load for an air source evap, they would be quick.

You would need to supplement the heating capacity in colder weather, probably with a backup gas or electric boiler side-connected into your primary loop. I've seen some fancy ideas with the heat pumps, but again, keep it simple and weigh the need against the typical hours expected to be below your limit.

My final observation is this system by nature would be far too complicated and would be best applied strictly for supplying all the chilled water required for the building while recovering all available heat for the hot water side. Based on my own experience with several large tonnage multi-source heat recovery systems many years ago, I'd stay away from the air source altogether. Just Run a cooling tower to reject you excess heat to a water cooled condenser in parallel with your heat recovery condenser.

[rjk_cmh]

The idea to vary the amount of refrigerant flow to the outdoor section comes from the same capability in VRF systems, such as City Multi systems, with the difference being that water would be used on the inside of the building instead of refrigerant everywhere. The idea is to squeeze that last bit of efficiency out, and not just be a little bit better. I see the following Pros and Cons:

Pros:

• A water leak is not the end of the world - just fix it and fill the loop. On a VRF system, a leak or failure in the line means that the entire system could go flat, requiring a recovery and recharge of potentially hundreds of lbs of refrigerant.
• 4-pipe air handlers could achieve supplemental dehumidification - VRF system "dry" modes usually just run the coil near freezing but don't have the option for reheat.
• No limitations of the amount of refrigerant charge in a system, since leaks would just be loop water and a refrigerant leak would be in the mechanical space or outside.

Cons:

• A leak would result in nasty loop water everywhere - a refrigerant leak would dissipate with minimal mess (maybe some oil near the leak).
• The refrigeration side would be quite a bit more complex than many systems (although still not as complex as most VRF systems).

When you say hot-gas defrost, that is what I was thinking, but I was concerned that the liquid produced from condensing the gas would need to be evaporated somewhere, causing an uncontrolled cooling effect, thus the need for a chill-water buffer tank. However, maybe an appropriately-sized accumulator is all I need if the defrost cycle will be relatively short.

Yes, it makes most sense to vary the compressor capacity based on deviation from setpoint of whichever water loop requires the most capacity, whether it be the cold loop or the hot loop. The idea was to size the chiller for the cooling demand, and be able to run it as a straight heat pump in the winter with supplemental boilers. However, I could maintain a chilled water loop even in the winter in order to run water-cooled refrigeration equipment like ice makers and freezers. If several ice makers are running, I could reduce the amount of refrigerant going to the outdoor coil in the winter, without banging back and forth with a solenoid.

It seems to me that it would be more effective and useful on smaller systems - the larger systems would have enough thermal momentum because of the water loop size. I was actually considering its use in large, high-end residential settings, not commercial buildings.

[icemeister]

I still believe it's better to keep it simple.

If you need heat from an air source, then I would build a separate air-to-water heat pump package and pipe it into the HW primary loop alongside the backup boiler. It would be a lot less headache, cheaper and probably more efficient. A straight HR chiller package isn't high tech, but the VFD would be a big plus. I were going to be responsible for it, I'd use the KISS principle and forget the integrated air source altogether. It isn't worth the hassle.

Just FYI...

The condensate from the HG defrost on a smaller system is typical handled by a suction/HG heat exchanger or as you say a water heat source (this is what Kramer Thermo-Bank HG defrost system is based on.). You'd also need a suction accumulator, typically with a liquid boilout coil.

I had one on an air source job with a 500 HP screw that had two 120 tons air cooled condensers circuited, for either evaporator or condenser operation. That accumulator was 36" diameter and about 9 ft high. We called it a SuperPot. The defrost was and alternating hot gas, where one unit stayed running as an evaporator while the other was defrosting. When one went into defrost, the whole neighborhood was calling the fire department because of all the steam when the fan kicked back on. LOL. The defrost was pretty fast though, at around 5 minutes, so any heating capacity loss was never really noticeable on the building's HW side.

[icemeister]

...If you need heat from an air source, then I would build a separate air-to-water heat pump package and pipe it into the HW primary loop alongside the backup boiler. It would be a lot less headache, cheaper and probably more efficient....

I had a second thought on this. It would be much more efficient heating to have the air-to-water heat pump in the chilled water primary loop dumping in 50-60 Deg F water to provide additional load for the HR chiller.

[rjk_cmh]

I had a second thought on this. It would be much more efficient heating to have the air-to-water heat pump in the chilled water primary loop dumping in 50-60 Deg F water to provide additional load for the HR chiller.

That's interesting, so it would act as sort of a cascade? Unfortunately, I have zero formal engineering training, so I don't know how to run the numbers on it. So the air-to-water would be piped in so that it would heat up the chill water loop when it was running, and then the HR heat pump would be more efficient because it had a load to work against - is that the idea? If so, that's even better - because I was considering using a CO2-based air-to-water heat pump, which requires low EWT according to the manufacturer.