OT: Ice thickness sensing in ice bank

[steveb1475]

Dan,

Based on the linked info I don't think its an open tank, it appears to be a heat exchanger where ice forms on the tubes but doesn't freeze solid. Process water flows through it so it can't be allowed to freeze solid. Maybe Steve can elaborate.

Yes, this is how it works. Ice builds up until a single point thickness probe is made. A contact from the probe would close the ammonia vapor regulator to stop production of ice.

BTW Steve, just out of curiosity, is the absorption/distillation driven by a burner, electric element, or does the process recover heat to drive it? Just curious because I think that there is a future for absorption chillers as we go green.

This isn't an absorption chiller, it is a basic ammonia vapor compression system.

DeltaT is going to relate to the load on the ice bank. How fast it is being depleted, not how much is left.

To use the electrical analogy, the OP is asking how much charge is left in the battery. Not how fast is it discharging.

Exactly.

Assuming you know the initial level of charge then the total Ah taken out will give you remaining Ah thus time at this rate. IF you are in doubt you crawled in the battery well with a hydrometer and measured the pilot cells.

The problem is that I never know the initial level of "charge", just seems a lot easier to find a sensor than to try to calculate ice production and usage.

 

[TConnolly]

I wasn't familiar with ammonia compression refrigeration and just assumed that it was an absorption chiller - that's what I get for assuming. After all this time you'd think I should know better.

I did some Google-ing about ammonia compression refrigeration and found some interesting information, including information on why the system is laid out as in your photograph (health and safety reasons).

Since the system uses a compressed gas cycle then the refrigeration loop performance may be a lot easier to determine - delta p across the evaporator may be all you need to measure as the pressure/temperature relationships of refrigerants are well understood. The compressor or evaporator manufacturer should be able to give you those tables.

You may need to perform some tests to determine just how much latent heat capacity you have when the single point thickness probe is made. From there you can determine the amount of heat put back into the chilled water by the process and the amount of heat taken out of the chilled water by the refrigerant. That should give you a good picture of how much capacity is left.

Sounds like an interesting/challenging project.

 

[Lancie1]

Yes, this is how it works. Ice builds up until a single point thickness probe is made. A contact from the probe would close the ammonia vapor regulator to stop production of ice.

Steve, Do you really care about the thickness? Is not the overall volume of ice just as important, or would serve as a substitue?

If so, install a small water overflow tank. As ice freezes it expands, and the volume of water forced out of the tank will be exactly proportional to the volume of ice formed. Measure or weigh the water in the overflow tank, and set up a scaling equation to then calculate the volume of ice on ALL the pipes.

 

[mellis]

Exactly.

OK, at least we agree on the problem.

Not sure I have a solution for you, but you can contact one of the radar level transmitter manufacturers and see if they think they can do it. Some are: Vega, Siemens, Rosemount. I've used them for liquid level, not exactly this application. It all comes down to picking up a strong enough echo from the ice surface.

Something just occured to me. What technology does the point sensor use?

 

[Gil47]

My comment come from my observations dealing with ammonia refrigeration systems

Ice forming on ammonia pipes starts acting like insulation would on the pipes, thereby reducing the ability to transfer heat, in turn reducing the amount of gas returning to the compressor, hence a slight reduction in the pressure, which causes the compressor to start unloading, causing the ammeter load to reduce.

The opposite effect occurs as the built up ice layers are being melted.

If your unit is a total self contained unit these changes will be easy to monitor.

Another thought that comes to mind is that it may be possible to monitor a light reflection or colour shade change as the ice level builds up, as colder ice or thicker ice forms, although this observation is most noticeable in the air, it could be possible to monitor for this effect where the suction pipe just leaves the water tank.

My thought is a light shining at the pipes and a light meter monitoring the reflcted light level.

Its an interesting topic thanks for bringing it up.

 

[leitmotif]

Your illustration of the heat exchanger in which ice is formed remind me of what I believe are called freeze plates.

Regardless of what they are called I am not sure how you are going to get a sensor iside that senses only the total volume of ice formed "inside the box".

As previous message asked what about the expansion valve sensor - is there some way you can take advantage of existing sensor and at minimum use it to tell you when you have maximum ice? At least this way as you stated with the battery example you would know your starting number.

I think watching return and supply temperatures is still the best way at least for interim. Supply temp will be proportional to area of ice (heat exchange surface) which is proportinal to volume.
So at 100% volume you have 100% area and minimal return temp ie 100% cooling from ice. In theory same at 75,50 and 25%. The one snag to this logic I think is volume is a cubic function and area is a square function.

Yes I understand startup and shutdown of large chillers is a bit more complicated than turning a light on or off. I also understand your dilemna with a too small unit and a too large unit.

Dan Bentler
Dan Bentler