2 Pipe - Heating to Cooling Tranistion 1

[KoolTemp]

I have a very interesting case concerning a 2 pipe system. The owner wants to transition from heating in the morning (boilers heating the supply water up to 120F) to coiling in the afternoon (chillers supplying 45F water). With this being a 2 pipe system I need to dump all the BTU's in the water medium (huge energy waste, but this is what the owner is set on) to a low enough temp where the chillers don't blow their refrigerant. On top of this there is a very tight time constraint of 1/2hr-1hr. There is only (1) small MUA available and this doesn't have near the capacity to get ride of this heat. There is also a pool but it is already tempered. A heat exchanger for the DHW was considered but by the time the heat dump is needed the peak demand hours are far from over for the DHW. The systems worst case gpm is 1125 gpm.

Any suggestions on how to control/dump the heat/ or make this system work properly? Any ideas at all?

 

[MintJulep]

Does the owner not pay the utility bills?

You need to put your argument against it in terms of operating dollars.

Then, if (or when) you still need to do this I'd think about two great big insulated storage tanks, one for hot and one for cold. Just "swap" the hot stuff and the cold stuff in your allocated 1/2 hour.

 

[DRWeig]

We've done that at a number of elementary / junior high / high schools before. It worked OK, but it wasn't fast in the larger systems.

There are two valves and a tee arranged like one big mixing valve in the return line from the secondary loop, feeding the water either to the boiler or the chiller. When the heat goes off and the secondary loop's water is hot, the three-way valve modulates to maintain a maximum temperature being returned to the chiller. The chiller can go ahead and start when its return water is below the maximum. The rest of the return water is mixed with the chilled water supply and sent out to the loop. As the loop cools, the modulating mixing valves keep the chiller return temperature safe. When the mixing valve is finally completely open to from the loop to the chiller, you're done.

The mixing control loops need to be tuned well to avoid overshoot. Rupture disk blowouts aren't cheap.

Mint is right, it's an energy waster and we used to joke that it was an entropy machine, not a hydronic system. But the building owner demanded it. They didn't want to pay for those two other pipes.

It can be done, though. Turnaround times were fast in small systems of 40 to 60 tons. Larger ones could not be made to meet your 1/2 hour limit, though. One of the largest schools we worked took over 2 hours to fully change over on a moderate day. Your mileage will vary. The time to cool the loop with mixing will depend on the total amount of water in the system and the hot water temperature at the time that changeover starts.

One other hint from a similar job in which my company was not a player: if you have some buried, insulated chilled water pipe going to and from some outdoor units, be careful with electrical heat tracing inside the pipe insulation. Some specifiers don't know that "self-regulating" heat cable also needs a thermostat. Sure, the heat output goes down as it gets warmer, but it doesn't go to zero until it's very hot. What we all learned back then was that self-regulating heat trace inside very good pipe insulation and buried underground, can cook the water up to 150°F or better before it reaches equilibrium. My company came in for the clean-up. The original mechanical contractor had to buy some compressors.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[DRWeig]

Thinking more about it, MintJulep's suggestion about two tanks in which to dump the hot or cold water sounds really good to me. It might get expensive if your systems are big and the water volume is large, that seems like the only possible disadvantage.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[cdxx139]

Depending on your code, IECC requires a 4 hour deadband when you change from heating to cooling or vice versa on a 2 pipe system. There is also a 15 OA degree deadband (like OA from 62-77), where neither heating nor cooling can happen.

Clients love the savings, but whens is 76 outside and there isnt any cooling, they come calling. It has gotten so bad, that we provide a memo explaining a 2-pipe system before we begin design, so they understand it exactly.

If you utilize an aggressive OA reset on hot and chilled water setpoints, then the temperatures will be closer to what you need, after the deadband is over.

Not the answer your looking for, but unless this is a process load, I find it crazy to expect a 1 hour turnaround, and potentially against code.



knowledge is power

 

[logmode]

You need to set up a reset on discharge boiler water from a two input, return water, and outside air. Then the hard part is to set it up so that before the chiller starts loads have used the heat and the water is cool.
You are the tec! The owner needs help. 120* is bs
logmode

 

[KoolTemp]

Thanks guys for all the input. I like the two tank idea and will look into this further. I have brought this being an energy waste up before but the owner still has this idea of this working and being the best idea in their head.

cdxx139 - can you explain this deadband issue a little further? I don't quiet follow you.

Thanks again

 

[MintJulep]

I have brought this being an energy waste up before but the owner still has this idea of this working and being the best idea in their head.

Then you have not done a good job explaining.

Put it in dollars. Capital and operating.

 

[cdxx139]

Kool,

IECC requires a 2 pipe system to have a 15 deg OA deadband and 4 hour changeover time.

No heating or cooling can happen (with the hydronic 2-pipe system between the 15 degrees you designate). So lets say between OA temps 63-77 degF. At 62 and below you provide heating hot water. When OA reaches 63, you stop heating water. When OA reaxches 78 degF, you can begin generating chilled water (provided a 4 hour waiting period has passed, since your boilers were energized).

This 4 hour period is to give time for the system water temperature to get closer to equillibrium. If you are aggressive in your OA temp reset, and size your coils for for low temp heating and higher than usual temp cooling (while still providing dehumidifcation), then this minimizes the energy loss between the changeover of system.

My system was a 2-pipe FCU system in a dormitory. I sized the FCU coils with the highest possible CHWS temp, that met load. From memory, something like 48 degF. (instead of the traditional 45 degF). Since it was a 2-pipe system, I reverse engineered the hot water coils, with the same GPM as cooling, and the lowest HWS temp, which was something like 110 degF. The cooling requiered a higher GPM flow, so I started with worst case (whcih was cooling).

Add OA reset, the HWS was 120 degF at 0degF, and dropped to something like 90 degF HWS at 60 degF. That allowed the HWR to be in the 70's-80's degF, when the deadband went into effect. After running the pupms to help dissipate the heat, and the 4 hour deadband, the HW is as close as possible to the CHWR temp, to begin the CHW system, without freezing the chiller refrigerant coil.

(Sorry for the length) A way of getting around this is to make the AHU's providing ventilation have DX coils instead of hydronic coils, so you can provide some cooling (or at least dehumidification) to the spaces.

knowledge is power

 

[MintJulep]

...many installations of this type waste energy and have been unsatisfactory in climates where frequent changover is required, and where interior loads require cooling and exterior spaces simultaneously require heat.

Changeover should be designed such that the chiller evaporator is not exposed to damaging high water temperatures and the boiler is not subjected to damaging low water temperatures. To accommodate these limiting requirements, changeover of a system from one mode to the other requires considerable time. If rapid load swings are anticipated, a two-pipe system should not be selected, although it is the least costly of the three options.