Reefer System Design

[Dale Tanski]

Pete,
Yea, they will leak from the atmosphere side if you pull a vacuum and they are separated from their mated half. 
Just a thought... if you perfect the system at home (no leaks) and all you do is cut the high and low pressure lines in one spot you only have two joints to get right on the boat.  You are better off pulling a vacuum on the complete system and charging at that point in real conditions.  Things change like air flows and temperatures.
Another point about leaks. You may think that you have found and repaired them all.  I have seen leaks that take 3 to 4 years in a small running system to drop the charge.  Because the system is not exposed to atmosphere (moisture) a 5 minute recharge and one is good to go for 3 to 4 more years.  Besides being good practice it just maybe a reality.  Good luck and congratulations refrigeration tech!
Dale

[PeteW]

Dale,
I should be snowing any time now in Buffalo but since its still over 100 degrees F out here in AZ, I found I only need a small amount of refrigerant to develop a vapor pressure of 100 psi. That  and some soapy water I was able to find the smaller leaks. I put shrader valves on the evaporator side figuring I would be breaking the seal on just that side. Sounds like a dummy set of Aeroquips would allow me to pump down that half, charge it and then plug it into the condenser unit. It was my inner environmentalist not wanting to dump a bunch of R134a into the ozone layer for no good reason. But a set of connectors is $70. A can of 134A is $20.

My bigger issue will be with the design in general. I now realize that a large holdover plate will take a very long time to freeze running a single BD50. Essentially you end up transferring all the energy stored in you batteries into the chill plate. You will need shore power or a runnable source to get the box chilled down initially.  An Evaporator on the other hand has little mass and creates cold air immediately. But your compressor will cycle on an off all the time. But I got it running today. So time to check my superheat, add some more refrigerant and tweak the expansion valve.

I am curious how long it will take to cool 30 lbs of eutectic fluid down to -5 degrees C from a 30 degree start temp. The phase change is enormous. Maintaining a 4 cu ft freezer and a 7 cu ft refrigerator I estimate at 9,000 BTU per day. The BD50 will pump 660 BTU/hr. So it will be running at least half the time if it ever even gets to -5.


Yeah, refrigeration mechanic.  A week ago I couldn't spell refrigeration. Just another boat project I somehow thought would be fun.  Thanks for your support.  Pete

[Dale Tanski]

Pete,

1) Usually cold enough around here to snow by the 2nd week in October.  We occasionally see some snow in the upper elevations 1200ft+ around that time. Certainly the leaves are gone.
2) Yes the vapor pressure at 100F would be impressive.  When you get it in the boat things will change. You can use compressed air if you have the pressure or any inert gas such as CO2 or Argon seeing as you are going to pull a vacuum anyway.  Compressed air depending how much moisture is involved could harm your dryer but the vacuum should pull that away.
3) They do sell dummy plugs for the Areoquips.  Many systems such as the cold machines are shipped with them.
4) Your inner environmentalist needs to be aware that millions of dollars of perfectly fine running systems were scrapped just because they changed from R12 and R22.  The Europeans are phasing out 134a starting in 2011 and gone by 2017.  California is talking about banning 134a already.  The amount of pollution generated by these changes will be far worse than leaving things alone.  The funny thing is the government makes more money via taxes from the sales of new replacement systems that those that are already running just fine.  I know... its is for the children...
5) As for the design, you are correct and semi-correct.  With the holdover plate you are front loading the energy into cold btu's.  With the evaporator plate you are going to add the same amount of energy to get the same amount of cold btu's.  That being said, lots depends on loss or should we say gain from the insulation surrounding the conditioned space, but theoretically, each method sees the same loss however the delta temperature is different with the difference in the time time it takes to change the temperature of the box using each method.
6) The hold plate design in a boat is best suited for an engine driven compressor set up.  This of course takes no electrical energy from your batteries, and we have more shaft horsepower than we can use for the prop.  An engine driven system with a hold over plate stores the cold energy in the cold plate from motoring out to your sailing area, charging your batteries and so on.  Theory of course must eventually meet reality, just like our current economy, and things just don't work like the text books say.  The problem is that if the plate does not pull all the way down (did not run the engine long enough) any supplementary electric system with a plate evaporator will work its butt off because that unsatisfied holding plate will suck up every cold BTU it can get. The holding plate essentially expands the capacity of your box without adding any interior space.  If you have the power it is not a problem.  Anything you place in the box once it is down to temperature does the same thing.  If you had a 25 pound frozen turkey, that acts just like a holding plate. The only difference is the phase change aspect but you could say the 25 pound turkey is a 100 pound chunk of meat with the phase change formula rolled in.
There is no perfect answer or design.  Just like life, every which way you go there is a series of compromises and assumptions.  If the assumptions are correct and you are willing to live with the compromises, life is good.  The key here is you are custom building a refrigeration system for your specific needs.  If your needs change you now have the ability to modify the systems design to meet the new requirements.
7) Remember the colder you go, the higher the delta T between the interior and the exterior of the box.  As that goes higher the higher the loss and the higher the energy requirements just to maintain.  One desirable design concept is to maintain the box at the warmest possible workable temperature.  This minimizes loss and minimizes the energy required to maintain the temperature.  -5C is really getting down there.  You are not storing food for months and I doubt that ice cream is the majority of your stores.  Ice cream needs to be colder than most frozen food because of the high sugar content by the way.  If you load frozen stores in a 40 degree box it will take days for that stuff to "phase change" minimizing your energy requirements. Unless you are planing on using the boat as the dooms day bunker think shorter term storage.  If we are talking bunker here, count me in and Email me under separate cover and let's talk.

Dale

[PeteW]

Refrigeration test data looks acceptable.  I was able able to freeze a 31 lb hold over plate in 95 degree F ambient temperature. I wrapped a few towels around it for insulation.  Here's the system I put together from old junk I bought on Ebay.




The pieces that I salvaged out of my 365 Ketch include the air cooled condenser, the base plate, 2 sets of Aeroquip connectors and the Danfoss controller. I upgraded to a Danfoss BD50 compressor from a BD35. The holdover plate was a steal and the thing is huge hence the horizontal mounting. Notice the 2nd capped off circuit that can be connected to a engine driven compressor at a later date. Also note the Danfoss TXN thermostatic expansion valve. This unit is adjustable over the range of +/- 40F.  This part is essential especially if you know nothing about designing refrigeration systems like I don't. FYI turn the adjuster CW to lower the steady state temp measure near the sensing bulb which is  strapped to the output suction line. It provides the feedback to this control element which dynamically meters the liquid compressed refrigerant into the evaporator coils.

A receiver (black vertical tube) is nice to have as it stores a  reservoir of liquid refrigerant.  If you have any small leaks you will not go low on refrigerant so fast.  Also a humidity sensing sight glass is a nice feature. The filter/drier and the receiver are new as they contain desiccant.

My question was , how long will this thing take to chill off and freeze? I have the data and have also done the math to back it up. From my data I determined that the BD50 pumps about 350 BTU/hr out of my holdover plate. The data sheet says it pumps closer to 500 BTU/hr at half speed so some significant amount gets lost somewhere. As expected.  Doing the physics, a 31 lb plate will take 31 BTU per degree F. So to cool from 78 down to 32 degree's will require 4 hours at 350 BTU/hr to remove the 1460 BTUs. That's what it took per my test data. At that rate, to freeze this plate however requires an additional 143.7 BTU/LBS to turn it to ice or 4455 BTUs total. (latent heat of crystallization). At 350 BTU/hr it will take 12 hrs to freeze this plate. I ran it for 6 hours and the plate was better than half ice. It froze over from the inlet, or in my case the bottom up. On the compressor I put a 470 ohm resistor in series with the thermostat switch to increase the RPM up to 2800 RPM from 2000. This jacks the current up and also make the controller run hotter. I put a small fan on mine. That rpm is  where the 500 BTU/hr guesstimate comes from. Suction temp went to 23 degrees F.

Superheat was consistent at 2 to 5 degress F. Subcooling also remained low, around 2 degrees. Some says that low subcooling is a measure of high efficiency. My suction line did ice up toward the end of the test. Some additional work on raising superheat may be needed.

It was a fun science project and it saved me some money.

Pete,  Pearson 365 Ketch #6

[Dale Tanski]

Pete,
Looks good to me.
Your cold (frost) suction line at the compressor is good in that it provides some cooling for the compressor.  You must be carefull not to allow sludging which is liquid both freon and oil returning back to the compressor.  This can be taken care of with a low side accumulator which would be good insurance for your expensive compressor and even more desirable on a cold plate installation where the unit tends to sit idle for longer periods of time collecting pockets of liquid.
You could wrap your discharge and suction lines together for a small section this will increase the suction temperature also minimizing the problem while lowering your high side temperature some.
Dale

[PeteW]

Dale,

What do you think about putting a section of evaporator coil in series with the suction side out of the holdover plate. It would help evaporate any residual liquid refrigerant and provide quicker ambient cooling in the box. I was thinking at minimum snaking the 3/8" return the long way out of the box through what would be the refrigerator side or solder some copper sheet to a section to get some surface area in there.

I could add a suction side accumulator to protect the compressor but I think my problem is that  I cranked the expansion valve open too much because  I wanted to go cold fast and then I turned the compressor speed up to boot. Maybe I need to slow the compressor back down when the unit approaches the set temp. That would meter the back the amount of refrigerant and discourage icing at the compressor. ( but my gut tells me that the expansion valve will simply stall and shut off )  Plus that will also requires some custom circuitry to be added to the digital thermostat. Or I'll simply keep it set at 40 F like you say. Frozen beer serves absolutely no purpose.

Unrelated, I always liked the idea that I could go anywhere in the world in what, once you have left international waters, is my personal floating sovereign country of which I am the ruler and the survival of the boat and crew is your highest priority.  Some might call that a bunker.  Ha!     Pete

[Dale Tanski]

Pete,
Any type of evaporator device will improve liquid flash off as long as it has no moving parts. It certainly can not hurt. 
The accumulator idea was just in case you were still in the science experiment mode. 
The throttle back of the compressor output as the plate reaches the desired temperature is a great idea.  Your concern about the expansion valve closing down is also valid however and that is why in a small system like the one you have cap tubes are so nice.  They have no moving parts.  This of course is their main disadvantage as well as flow adjustment is harder to establish.  Your TEV is easy to adjust and more efficient over large temperature variations but with the exception of initial start up you box really does not have large temp load swings.  Cap tubes would be a great match for a variable speed compressor set up. 
Dale

[PeteW]

Here's some pictures of my refrigeration system after it was installed in the Pearson 365.



The holdover plate is 31 lbs and was too large to fit on the narrow wall. This one is was a new old stock item  I found on Ebay.



Here is the home-brew condenser unit mounted on a glassed in shelf in the port sail locker. The compressor is a BD-50 (new) but all the other parts except the fan, filter drier, site glass and receiver used. The little aluminum box contains a 12 volt switching regulator module that I bought from Digi-key electronics. It allows me to back off the 200 CFM 12 volt fan efficiently with a potentiometer speed control (see black knob).

The test results showed promise. From a standing start after 6 hours of initial run time the plate got down to 14 degrees F and the ambient air center of the box was 31.8 degrees F. In 1 hour the eutectic in the plate dropped from 52 to 32 degrees F. Doing the math: 20 degrees X 31 lbs = 620 BTU's/Hr capacity. 30 BTU less than the rated capacity of the BD50 running at 2800 RPM.

With the plate frozen over the box  remained cold until the next day.

Pete W   Pearson 365 Hull#6