Adventures in Refrigeration

Refrigerators in boats aren't typically like those at home.  Unless you have an endless supply of power, you need one that is a bit more power efficient.  In my case, the refrigerator cabinet is a large, fairly well insulated cabinet that opens at the top (like a good portable cooler) and has a divider inside that splits the space into a freezer section and refrigerator section.  The "cold" in the freezer is provided by what is called a cold plate system. There are two metal plates filled with a liquid that will freeze when the refrigeration compressor runs and then they slowly absorb the heat.  Think of it kind of like a block of dry ice that is repeatedly re-frozen to keep the freezer cool.  The refrigerator side of the cabinet is kept cool by allowing some of the heat from the refrigerator side to passively bleed over into the freezer through a couple holes in the divider between the compartments.  

Freezer and Refrigerator Controls

This works OK,  but has a couple issues. The temperature of the freezer plays a large part in the temperature of the refrigerator.  Both the refrigerator and freezer are colder at the bottom than at the top.  If you open the refrigerator much, the refrigerator side can warm above safe temperatures and take a while for the spillover air to cool it back off.  Conversely, if you don't open it enough, things can freeze in the refrigerator side, particularly at the bottom.  And while I like a cold beer, frozen ones can  be an issue. And finally, my freezer could only reach a low of around 15° F (ideal is near 0°F).

One more issue, specific to my fridge, is that the genius who installed it placed the expansion valve (the point in the refrigeration circuit where the cold starts to be generated) outside of the refrigerator cabinet.  This valve freezes and thaws with each refrigeration cycle and therefore condenses water out of the air and makes a big wet mess in the cabinet.  

Other than some of the issues above, my refrigerator did work OK, until the last few trips.  The refrigeration compressor started developing an issue where it would try but fail to successfully start up until after a random and increasing number of attempts. This caused the freezer to warm to unacceptable levels at times. After some investigation, I could only narrow it down to either a failure of the controller or the compressor.  Since those components make up the majority of the system, the system is over 30 years old, it was showing a fair amount of corrosion, and I wanted to correct that installation problem, I decided to bite the bullet and replace the entire system myself.

The old cold plate configuration

This may seem like an excessive step to take, but new systems come pre-charged with refrigerant and special connectors so they can be done as a DIY projects.  Given the difficulty I've had in the past hiring "professionals" to come work on the boat, I actually thought I will either break even or come out ahead...and I wouldn't have to be at the mercy of a "professional's" schedule. So I did a little research and decided I would replace the refrigerator with a newer version from the same manufacturer (SeaFrost).  A similar system should at least minimize the complications of the install.

Replacement System
(image from SeaFrost)

Of course, as often happens with projects, what started out as a simple swap-out of the old system for a new one and the relocation of the valve location slowly grew with several of those "while I'm doing this, I might as well..." thoughts.  The freezer would freeze, but not quite cold enough for ice cream.  Ok, actually, the colder the better for all longer term food storage...but ice cream would be nice too.  When talking with the SeaFrost salesman, it didn't take much convincing for me to upgrade from the smaller equivalent replacement to the next step up with a bit higher capacity.  The upgrade also had the ability to control compressor speed, so I could turn it down and and achieve the same results as the smaller system while using less power.  The existing system had the cold plates mounted in an L shape along adjacent walls, and the salesman highly recommended moving one plate so the two would be on opposing walls for more even freezer temperatures (stuff in the freezer would be sandwiched between the plates).

The old fridge removed

Naturally, if I could get the freezer colder, this would likely exacerbate the problem of the beer freezing in the bottom of the refrigerator.  So I began to think about how to solve that issue.  I had heard of others using small fans to mix up the air so the cold wouldn't settle at the bottom. I also thought of adding some sort of movable shutter to manually control how fast the "spillover" between the freezer and refrigerator would occur.

Patching holes and installing
a new cold plate

I finally settled on a theory that wasn't too mechanically complex and I believed would solve several issues.  My theory was that I would plug the two holes at the bottom corners of the divider, I would construct a small duct on the refrigerator side of the divider that would route the air from the hole about 1/4 of the way up the wall to near the top of the refrigerated section, and I would put a fan at the top of that duct that would pull cold air from the freezer and dump it out near the top of the refrigerator. I would leave the gap between the refrigerator lid and the top of the divider alone to allow return airflow.  Since warm air rises and cold air falls, I figured this duct would limit the majority of the spillover without a mechanical shutter to close the hole.  Then, when the fan was turned on, it would draw the cold air from the freezer up the duct and spill it over the top of the refrigerator.  As the cold air fell down through the refrigerator compartment, the warmer air of the fridge should rise and the pressure differential caused by the fan would cause that warm air to enter the top of the freezer section to be re-cooled.  This mixing should also make the temperature more consistent from the top to the bottom of the refrigerator. And by drawing the cold air from the higher point in the freezer than the very bottom, there would be a section at the very bottom of the freezer where the coldest air would still pool and become a deep freeze section that wouldn't be impacted by the bleeding off of cold air for the refrigerator.

New plates installed with valve
inside refrigerator box

I'm no thermodynamics expert, but this all sounded reasonable to me. So I decided to give it a try.  I installed the new refrigerator...which takes much more time than that one sentence gives it credit for.  Carefully bending copper tubing so it doesn't kink, by yourself, while working in cabinets where you can barely reach what you are working on is not trivial.  Then I found a vinyl (PVC) downspout at the local home improvement store that I could cut into a U shape and create a duct.  Estimating that the refrigerator section was around 5 cubic feet, I ordered a fan that would move about 2 cubic feet of air per minute.  This would theoretically do a complete air exchange in the empty cabinet in about 2.5 minutes.  I figured there would be a balance that would need to be figured out between taking cold air from the freezer too quickly and cooling the refrigerator too slowly, so this was my best guess at a reasonable flow rate.

Testing the duct

In order to experiment a bit, I temporarily sealed the lower corner holes (and the wire/line access hole) with excess packing materials, built the duct using some PVC glue, and ordered a small dual-bearing computer fan and a digital temperature controller for it.  I ran experiments and found that I did seem to have positive control over the refrigerator temperature.  With the freezer at about 15° F at the time of the test, I set the other controller to turn the fan on at 37°F and off at 36°F. The temperature in the refrigerator would slowly fall when the fan was on (about 30 minutes to go from 37°F to 36°F) and would slowly rise when it was off (about an hour to rise from 36°F to 37°F).  I also checked the temperature at several levels within the refrigerator compartment and found only 1 degree of difference between the top and bottom.  During the tests, I did not observe any noticeable rises in the freezer temperature beyond the normal operation.  I did run the tests with both the fridge empty and with several milk jugs filled with water to simulate a partial refrigerator load.

Fridge duct with fan installed

Happy with the results, I made the changes permanent.  I did make one additional change and lined the lower half of the duct with some thin, foil backed, closed cell foam insulation to help further reduce unwanted bleed over.  After warming the refrigerator cabinet back up, I attached and sealed the duct with silicone, cut foam to make removable seals for the bottom corner holes, ran the wires and temperature probe needed to control the fan, and sealed the cabinet access.  I added a magnetic switch in the fan circuit so the fan would shut off when the lid was opened, and I installed a small electrical project box to hide the connections and extra wire from the fan and switches.  

So far I'm pretty happy with the results.  The only thing I'm wondering right now is if I need to increase the size of the fan to improve the speed at which the refrigerator section cools.  But I think I need a bit more experience with the current behavior before I make any changes.  I can also probably tweak the settings of the temperature controllers for both the refrigerator and freezer sections; however, this will be done slowly over time as I determine a sweet spot once the whole system stabilizes.  The first picture above is the system running with the freezer and refrigerator cabinet temperatures displayed.

I'd also like to give a shout-out to the folks at SeaFrost.  They were very helpful with advice and design recommendations on this project.  I wish more marine industry parts and service suppliers were this helpful and responsive.

Now I have a reliable refrigerator/freezer that acts a bit more like a home system than it does the normal boat cold plate system.  

A Couple of Projects

 Well, if you have read through my blog, you probably have a good idea of the amount of maintenance required to keep a boat going.  And this post will be no exception.  It still amazes me how many new issues appear when you leave a boat sitting for even a short period of time.  So, upon returning to the boat I had a list of minor things to address (for example, some running rigging needed to be replaced and some leaks needed to be addressed) and a couple of bigger projects.

The first project was the inverter.  As I mentioned in my previous post, my old Xantrex Freedom Marine inverter decided to give up during our spring shakedown cruise.  I pulled the unit and did some troubleshooting and found a blown internal fuse.  I tried replacing it, however the fuse immediately blew when attempting to start it up...so something in the internal circuitry has obviously failed.  I guess I could try repairing the unit, but it is 30+ years old and there are newer options out there that provide pure sine wave output and are more efficient.  It made the most sense to replace.  Of course, I don't want to rewire the boat, so in addition to finding newer technology, I needed to find one that would fit in the available space.

I narrowed my search down to two or three options that would fit, were at least theoretically suitable for the marine environment, and would provide pure sine wave output.  Two of the options could be found listed on marine supplier websites with prices north of $1500.  The third option was from Renogy, a company that is more well known for solar and off-grid land based setups, but claimed their unit was marinized and could handle the more corrosive environment found on a boat.  The fact that their unit was also 1/2 to 1/3 the price of the others, I decided I would give it a try.

Renogy 3000w pure sine wave inverter
(Image from Renogy.com)

The unit was a bit longer than my old one, fortunately, that was the one dimension where I had a bit of extra space.  Otherwise, it was a direct replacement as it specified the same gauge battery cables and fuses and was a mostly drop-in replacement.  Wiring was pretty straight forward, and other than finding and replacing a poorly crimped cable end on one of my cables, the unit itself wasn't too difficult to install.  The unit comes with a remote panel, but unlike the Xantrex, it is not very informative, just a switch with a couple of status LEDs.  A remote that mimics the on-board display would be a better option, and hopefully they will consider making one.  The remote was also smaller than the Xantrex, so I had to custom fabricate a plate to mount the new remote where the old one was.  Fortunately, this was easy to accomplish by purchasing a 3-gang black switch plate from the local hardware store and making a few modifications.

Other than the remote display and a couple minor quirks, so far I really like the unit.  Unlike the Xantrex, it has a very clean, pure sine wave output. Things like my induction plate and microwave are happier about the supplied power.  It also switches over to inverter power much faster than the Xantrex, so the microwave clock doesn't need to be reset and the TV doesn't even flicker.  The most annoying thing about it is that the battery charger cannot be disabled when the unit is on and the AC circuits downstream of the inverter are not active in bypass mode when the inverter is off.  These are minor issues and easy to work-around, but would be nice features to add.  Given I could buy 2 or 3 of these for the price of the others, to me it is a worthwhile tradeoff.

Another bigger project is replacement of my old chartplotter.  The main reason for its demise is that the prior owner of the boat did quite a hack job when installing that resulted in a large hole behind it and so I never removed it from the helm.  I could possibly replace just the screen and maybe the seals that caused the current problem with the unit, but just like the inverter, the entire unit is older and I suspect other internals are likely soon to have problems as well.  I could buy a new unit from Garmin that would be mostly a drop in replacement with the same 8" display, but that will set me back about $2k...not to mention I'm not a huge fan of Garmin.  

The other option to replace the chartplotter is to essentially build my own.  There are a couple of open-source software packages out there that are capable of replacing an existing chartplotter and can better integrate most of my Garmin and Raymarine instruments.  For the cost of a Raspberry Pi computer, a touchscreen display, and a few other electronic bits, I could have a more capable chartplotter than what I have now.  All that for what should come in at under $500 and some of my time.

Openplotter running on my Raspberry Pi.

I decided to go this route and have been working on the solution for a little while.  I'm using OpenPlotter, which includes the OpenCPN chartplotter software, SignalK for data integration, and a variety of interfaces to deal with a variety of instrument data.  Above is a picture of my working system on a 15" touchscreen display that is temporarily hooked up to my boat instruments.  You can see current wind, depth, location, direction, speed, pitch and roll information.  You can even see the AIS data from a few nearby boats in the marina.  Not displayed but working is the radar information and overlay.  Since the computer isn't mounted yet, the sensors aren't fully calibrated so some of the information like pitch and roll are a bit off...but it is sufficient for proof of concept.  And for the first time since I've owned the boat, I can now send route data to the autopilot.  The only thing I haven't figured out yet is how to get more detailed sonar information as Openplotter doesn't seem to have a means of supporting a fish-finder type of sonar.  This solution will also be setup so I can easily detach the display and store it inside.

The final bigger project I'm working on is my refrigerator.  Over the past several years it had started developing an issue where the compressor made multiple attempts before it successfully started.  Originally, this was infrequent, but more recently it has happened with increasing frequency to the point that it was no longer reliable.  While looking into the issue I noted that the compressor was rather old and had significant surface corrosion.  Apparently the original controller had failed at some point and was jury-rigged with one from Alder Barbour.  So, my best guess is that either the controller or the compressor was the likely the current issue...of course, that makes up the vast majority of the system.  As a reliable refrigerator is necessary for any longer trip, I decided to bite the bullet and replace the whole system.  New systems have more efficient compressors and often come pre-charged so at least I won't need a refrigeration technician to come install it.

Seafrost BDxp Refrigerator system
(image from Seafrost)

The original system was from SeaFrost. Not wanting to significantly alter the setup, I decided to go with them again.  After talking with someone from SeaFrost (I think it may have been the owner), I decided to get the upgraded version with a bit higher capacity.  The belief is that it should be sufficient to store ice cream (something that was just a bit beyond the ability of the old system).  The old setup of the refrigerator is what is known as a spillover system.  Some of the cold of the freezer "spills over" into the refrigerator side through some holes to provide refrigeration.  While this does work, it does have some drawbacks.  There was a tendency for items at the bottom of the refrigerator section to freeze while the top was barely within the safe zone for a refrigerator.  With the increased capacity of the new system, I think this issue might become worse.  So, I also intend to alter the spillover system by adding a duct, small fan, and temperature control in hopes that it will better meter the temperature of the refrigerator as well as provide a bit of air circulation to keep the temperatures a bit more uniform.  

So, those are the bigger projects currently going on or completed.  Hopefully they all will be done soon and I can head south and to the Bahamas.