Lithium Bromide Absorption question

[John_Galt]

OK, things are a little dangerous here. I have my notepad out and I'm doodling, and my thoughts are wandering to refrigeration and air conditioning.

Here's my big question.

How much water (refrigerant) is evaporated from the solution in a Lithium Bromide absorption unit?

Or, in other words, what is the concentration of the "strong solution" and the "weak solution" in the system?

Here's my doodling. I have an application where I need about 20,000 to 24,000 BTUs of cooling. In this application, I have a nearly limitless supply of waste heat from a nearby process. This heat is in the form of fluid at a temperature in the 180F to 200F range (80C to 95C).

I'm thinking that this could be used as a heat source for a Lithium Bromide absorption system.

For the cooling capacity I need, the system will have to evaporate 2.5 to 3 gallons of water per hour (20 to 25 pounds of water, or 9 to 12 Kg if metric is your thing). Actually, I'm not 100% sure of that number, because I base that on BTU's required to boil water at 100°C. I think that number varies some with temperature and pressure, and I haven't looked up the actual number for the vacuum/temperature of evaporation in a Lithium Bromide absorption system. I hope I'm close on that number.

In a conventional system, it's easy to know what the flows will be and what size compressor you need. If you need to evaporate x pounds of R22 in an hour, you a compressor that will move x pounds of R22 in an hour.

With a Lithium Bromide absorption system, you use a "solution pump" to move the solution around. The problem is, the solution contains water that is used as the refrigerant, and more water (probably much more) which is used as a solvent medium for the Lithium Bromide dessicant.

So, I'm curious, if I need 20,000 to 24,000 BTU's of cooling, which equates to 2.5 to 3.0 gallons per hour of refrigerant flow, will I need 5 to 6 gallons per hour of solution flow in the systrem? Or 50 to 60 gallons per hour? Or 250 to 300 gallons per hour?

If half the water in the solution "boils out" as refrigerant, then 5 to 6 gallons per hour of solution flow delivers the required cooling capacity. If only 5% "boils out" as refrigerant, then I need 50 to 60 gallons per hour of solution flow. If only 1% "boils out" as refrigerant, then I need 250 to 300 gallons per hour.

None of the "how it works" diagrams that I've found on the internet show solution concentrations. They just have "diluted solution" and "concentrated solution" or "weak solution" and "strong solution." Without numbers, how can one design a system?

I'm still in the "doodling" stage on this, but if it it feasible, I could develop a prototype or custom system for this.

[marc5180]

Taken from a post on here a few years back;

Lithium Broide solution (60%) along with refrigerant (water) (40%) is in the bottom of the absorber.

As far as the simple circuits go there are normally 3 concentrations on a single stage absorber. Strong solution leaving the generator 63% to 64% concentration. Intermediate solution spraying over the absorber tubes 60% to 61% concentration, this is a combination of concentrated and dilute solution. Dilute solution leaving the absorber, 58% to 59% concentration.

Most single stage machines run a 5% solution spread between the dilute and concentrated solutions.

Concentrations vary by manufacturer so these are just some figures you need to be close to at full load with design water flows. It is kind of a balancing act, you need to absorb as much refrigerant as you are generating and visa versa.

[John_Galt]

Most single stage machines run a 5% solution spread between the dilute and concentrated solutions.

That's the tidbit I was looking for. Great answer. Thanks for taking the time to answer, and to provide so much good informtion.

So, at a 5% spread, to achieve 24,000 BTU/hour cooling, I'll need to pump 60 gallons per hour of the solution. That's a key piece of information.

Taken from a post on here a few years back;

Lithium Broide solution (60%) along with refrigerant (water) (40%) is in the bottom of the absorber.

As far as the simple circuits go there are normally 3 concentrations on a single stage absorber. Strong solution leaving the generator 63% to 64% concentration. Intermediate solution spraying over the absorber tubes 60% to 61% concentration, this is a combination of concentrated and dilute solution. Dilute solution leaving the absorber, 58% to 59% concentration.

Concentrations vary by manufacturer so these are just some figures you need to be close to at full load with design water flows. It is kind of a balancing act, you need to absorb as much refrigerant as you are generating and visa versa.

I noticed when I searched for "Lithium bromide solution" that most commercially available solutions are 54% LiBr by weight. That's actually very consistent with the 60% number you were quoting.

When the solution is introduced into an evacuated envrironment in side the absorption chiller, some of the water will "boil off" and form the atmosphere within the chiller unit. The vapor pressure will be very low because water vapor pressures at ambient temperatures are usually very, very low.

The fact that the solution is strengthened about 6 percent from the water lost to vaporization inside the evacuated environment tell me another key that I was going to ask next.

Actually, I have a few more questions.

The first question was going to be about charge density for the chiller.

Question: How many grams (or cc's) of solution are introduced into the system per liter of interior volume?

If the evaporation of water from the solution in the vacuum inside the unit raises the concentration of the solution from 54% to 60%, it's a matter of some math to determine how many grams (or cc's) of solution are introduced per liter of volume within the chiller. I'm not sure the exact calculations. I may need to do some further research on this.

Also, you mentioned having to balance absorption with generation. I'm guessing that this balance is achieved, in part, by the relative sizes of the generator/condenser vessel (or section) and the evaporator/absorber vessel. Every diagram I have seen shows the generator/condenser as being smaller in size than the evaporator/absorber.

Question: Is there an ideal ratio for the volumes of these two vessels (the generator/condensor and the evaporator/absorber)?

Also, it seems that some diagrams show these sections as being in two completely separate vessels. Other diagrams show them very close together, perhaps within the same vessel, with dividers of some kind between the sections.

It would seem to me that when actually building a machine, it would make more sense to have the four sections isolated in two spearate vessels, conected by piping, but with insulation between the two vessel containers to keep them thermally isolated as much as practical.

Considering these things, I don't think it's even a question. Separate vessels with thermal insulation separating them (except for the required plumbing connecting them) seems to be the better idea.

And my final question is about materials. I was considering building something entirely out of stainless steel. I have material available for this, including stainless plumbing, tubing for heat exchanger sections, and stainless sheet and structural tubing for the structure itself. Actually, I'm thinking this will be a good project for practicing my TIG skills.

My thinking is that using stainless will allow me to not worry about corrosion, or deal with corrosion inhibiting additives in the LiBr solution. The down side is that stainless steel is not as conductive to heat as other more conventional materials (especially copper tubing).

Question: Is stainless steel an appropriate material for all parts of this chiller unit? Or are more conventional materials better (copper, mild steel, etc), in spite of the needed additives for corrosion protection?

I guess another possibility is using copper tubing for the heat exchanger plumbing within a stainless steel structure. I'd still have the corrosion issues with the tubing, and need something to keep the LiBr solution from eating the copper tubing, or clogging it up, but the main structure would be less affected. Also, I haven't researched the chemistry for the additives, but I would only need protection for copper (and possibly for the solder alloys joining he copper to the stainless where the tubing goes through the walls of the vessel structure). I wouldn't necessarily need corrosion protection chemicals to protect mild steel in that type of design.

There would be some additional questions about this. Specifically, what type of solder would be best for joining the copper to the stainless where the tubing goes through the vessel walls. I have some past experience with cadmium-zink-silver solders (I think I even have several rods of KappTecZ™ - High Temperature Solder for Dissimilar Metals still around from a previous project).