Ammonia Chiller Plant: Glycol Buffer Tank vs Hot well/ Cold well Tanks

[Polisie]

Hi.

I am busy with a project where we use an ammonia chiller plant. Reason being that the COP is above 7 and power in South Africa is getting more expensive.

Background
The refrigeration system offered will operate on a gravity feed re-circulated liquid ammonia design philosophy, with all ancillary items i.e. plate heat exchanger, compressors, vessels and piping, designed for this application. The refrigeration system has been designed to operate at a suction condition of -1°C and a discharge of 35°C. The main ammonia plant design is based on a single plate heat exchanger undertaking the heat transfer from the water to the ammonia.
The secondary side of the plant is based on individual heat exchangers (Evaporators) undertaking the heat transfer between the glycol and the product/room respectively.

We are running two(2) Bitzer compressors where each unit has a capacity of ± 96 kWR @ -1°C / +35°C @ 1450 rpm.

Question
I do not have a lot of experience on buffer tanks and hotwell/coldwell tanks. We looked at a Buffer Tank but the design engineer is on leave and I am stuck with the decision.

If we go use the Buffer tank we need to fill/pressurise the tank with nitrogen so that the propylene glycol (25%) does not degrade due to oxidation.
We sized it to a Dia600(mm)x1800(mm) high and a maximum pressure of 600kPa. The system will be pressure tested to 600kPa.

1. Is the tank sized adequately?

2. How much nitrogen pressure is required?

3. Or is Nitrogen required?

4. Which is better: buffer tanks OR hotwell/coldwell tanks?

Chiller Plant - PHE.jpg

[HVACRsaurus]

We would call that an "expansion tank". For recirculated glycol with plate HX it's usually a expansion thank is used. Hot / cold tanks are generally only used on systems with varying chiller/s capacity & varying process demand.

As for the size, it's all about absorbing the thermal expansion & contraction of the glycol. I cannot quote values but the bigger the circuit volume, the larger the expansion tank. Using my "ready reckoner" the tank described is fairly large, i expect that should accomodate expansion / contraction of several thousand liters of glycol (at a guess). Perhaps others can clarify

[PaulZ]

Hi Polisie
If you are concerned about the degradation of the Glycol have you considered a closed loop system.
This does away with the buffer tank and has an expansion tank instead. This tank is generally made of fibreglass and has a large rubber bladder inside.
We have installed several systems of this type and they work well as long as they are set up correctlt in the begining.
Paul

[Polisie]

Thank you.

Yes it is a closed system. The total glycol volume in the pipe and coolers are 1250 litres. The tank was designed on 600kPa design pressure at a temp of -5C, content propylene glycol 25%.

What I want to know is that the "air" space in the tank, can it be replaced with nitrogen and to what pressure should it be pressurerised? Can the nitrogen pressure in the tank be 220kPa?

The pump specs are 18 m3/hr @ 40 m and 25% propylene Glycol at -4C.

[RANGER1]

Thank you.

Yes it is a closed system. The total glycol volume in the pipe and coolers are 1250 litres. The tank was designed on 600kPa design pressure at a temp of -5C, content propylene glycol 25%.

What I want to know is that the "air" space in the tank, can it be replaced with nitrogen and to what pressure should it be pressurerised? Can the nitrogen pressure in the tank be 220kPa?

The pump specs are 18 m3/hr @ 40 m and 25% propylene Glycol at -4C.

Polsie, no air space at all in expansion tank or glycol system.
Install on suction side of pump.
Expansion tank should come with insructions on pressures etc.
Everything contracts when cool, expands when warms up.
Glycol system could also have a relief valve installed for extra protection.

[tjk2007]

hi,

pls refer the following... this is from ASHRAE...

ETV=	Vs [(p1/p2)-1]
	1-(Rf + Vf)
where,
ETV=	Expansion tank volume		m3
VS =	system secondary coolant volume at t1 temperature. (your case 3°C )
p1=	density at t1 temperature		kg/m3
p2=	density at max.temperature			kg/m3
Rf =	residual volume of tank liquid (low level) at t1 , expressed as a fraction.
Vf =	volume of vapor space at highest tempeature, expressed as a fraction.
VS =	1.25		m3
p1=	1030		kg/m3
p2=	1020		kg/m3
Rf =	0.1		Assume as 10%
Vf =	0.2		Assume as 20%
ETV=	0.018		m3 only
As per drawing:	0.50868		m3

[Boomer333]

Yes you can replace the air space with nitrogen , I am curious tho , are you not using an inhibited glycol , & are you not able to test the contents of the glycol to insure that it is at the proper levels of whatever treatment you are using on the glycol???
The pressure in the tank will depend on the pump suction , you want to keep enough pressure on the system to stop the pumps from cavitation on the impellers , pressure will also depend on just how much heat transfer you are expecting to see on the return loop , which will translate to pressure.
Have used both bladder tanks , as well as exp tanks , prefer the regular style exp tanks as you can monitor the air volume in the tank , where , with a bladder you might not always know when a bladder ruptures , causing loss of the tank on the system.
You should also insure that the piping to the exp tank is large enough for the tank , never run a 1/2" line , & 3/4 max is about 8' or so , after that , refer to ASHRAE charts.

[MikeHolm]

From the drawing, this does not look like a bladder tank. If the static fill is only half the tank, the rest will be air (or nitrogen) but I wouldn't worry about having air in there. Any O2 in the air will move to ferris components and you will be left with mostly nitorgen and CO2 anyway (as long as there is no new air getting in). This is a standard closed loop heating system. The glycol is well within its range and should last a long time, given the temps.