When you have a constant supply of power, such as with a house fridge, the compressor can run whenever required. In this case, the compressor works with what is called an evaporator coil. The evaporator is cooled by the expansion of the refrigerant gas, and the evaporator cools the surrounding air. Sometimes a fan or two is involved to move air over the evaporator. Because the evaporator/compressor combination runs often, the load on it is light. The compressor is therefore smaller than what would be found on a boat.

A boat, without continuous power available, requires a compressor system that can be run for short periods of time, typically when the engine is running. This being the case, some form of storage for the cold created by this compressor is required. Here holding plates come into the equation.

How Holding Plates Work

Certain liquids go through a change of state when they are cooled below a given point. Large quantities of energy are required to achieve this change of state. Fresh water is a good example. At 32 degrees Fahrenheit (0 degrees Celsius), water begins to freeze. In the process of freezing, huge amounts of energy are absorbed, changing the water to ice, roughly 1,140 Btu per U.S. gallon (3.8 liters). Because it is a true eutectic solution, all of the water in a given area will freeze before the temperature starts to drop beyond the eutectic point. When the eutectic begins to thaw, the reverse happens. The same amount of energy is required to warm the solution up past the freezing point. The temperature of the solution will remain constant until the last molecule of the eutectic has had its change of state. This eutectic property is the key to storing cold energy in the fridge system. With a true brine solution (salt-based) at a temperature of around 7 degrees Fahrenheit/-14 degrees Celsius, you have a holdover capacity of approximately 800 Btu per U.S. gallon.

Here's a simple experiment you can conduct at home to prove this principle. Take a glass of water with several ice cubes, and check the temperature of the water. After a while the water temperature will be at 32 degrees Fahrenheit/ 0 degrees Celsius. It will remain at this temperature as the ice melts, as that is the eutectic point of fresh water. On your thermometer you will see a steady temperature until the last of the ice has melted, at which point the temperature of the water will begin to rise.

Unfortunately, there are few true eutectics. Fresh water, as you know, is one. Adding salt to water maintains the eutectic properties but lowers the temperature at which the freeze/thaw point is reached. Salt, however, is corrosive and cannot be used in plates unless the plates and plumbing are made of stainless or galvanized steel. You cannot use salt solutions with stainless and copper, for example, due to corrosion occurring in the copper plumbing.

Just to complicate this problem a little more, there is a mechanical problem with water-based eutectics. They expand as they freeze. This can cause all sorts of problems with your plate and plumbing, unless proper allowance has been made in advance.

Non-Eutectic Solutions

All sorts of other ingredients can be added to water to reduce the freezing points. One is ethylene glycol, the antifreeze you use in your car. Another approach is to use methylated spirits with water. Both of these lower the freeze point of the solution. However, because they are not a true eutectic solution, several things happen.

First, there is no steady temperature point at which the plate will sit on a stable basis for a long period of time. This makes it difficult to keep the fridge or freezer at a set temperature. Second, because there is no true change of state, there is a loss in holding capacity of anywhere from 40 to 65 percent compared to a true eutectic. Finally, because the temperature continues to drop as the solution is cooled (rather than sitting at a fixed eutectic point until the whole plate is frozen) the compressor operates at a lower average temperature. This reduces compressor efficiency.

Super Juice

Tim James and Arne Adams, two rocket-scientists-turned-cruisers, saw the problems with the two choices just discussed and set out to solve the problem. They came up with a patented true eutectic solution with a capacity to absorb heat similar to that of salt water, but one that is noncorrosive and that contracts rather than expands upon freezing.

We built Sundeer's fridge plates in the usual fashion in New Zealand Ñ a combination of stainless liners, copper tubing, and a water/methylated spirits solution.

However, we made allowances for changing the solution when we got the boat back to the States. Upon our return, Tim and Arne came down to the boat and connected a strip chart recorder to the freezer. We tracked the holdover, temperature ranges, and amp hours consumed on a daily basis.

We then changed the solution to their super juice and went through the same data again. Where before we had a 36-or-so hour holdover, we now had a capacity of four days plus. And, compressor amp hours (or running time) was down by 40 percent!

These results are typical for any true eutectic solution.

Eutectic Temperature

Whether you make your own plates or buy them already made, you need to decide what temperature eutectic solution to go for.

The industry standard is 26 degrees Fahrenheit for fridges and 0 degrees to Ð5 degrees Fahrenheit for freezers. The temperature at which the box will run is a function of the surface area of the plate, air flow over the plate, heat load on the box, and temperature of the frozen solution inside the plate.

Two issues push us toward the highest acceptable temperatures. One, the warmer the box is, the less difference there is between inside and outside air and the lower heat load is on the insulation. The lower the heat load, the less compressor running time.

If the ambient temperature is 75 degrees Fahrenheit, and the freezer is at 0, there's a differential of 75 degrees. If you let the freezer run at 15 degrees Fahrenheit, there is only a 60-degree differential. That's a difference of 20 percent in heat load.

The other issue is compressor efficiency. The higher the temperature at which it operates, the more efficient it will be.

Surface Area versus Temperature

We mentioned plate surface area as having an impact on temperature. For a given heat load and given box size, you need a certain plate temperature to achieve your desired results. By increasing the surface area of the plates you can achieve a lower temperature in the box. Alternatively, you can use greater plate area to maintain a given temperature while raising the eutectic temperature (which means a more efficient compressor).

Most stand-alone plates are held off the surface of the fridge or freezer by about 3/4 inch (19mm), so that air can circulate around both sides.

This helps plate efficiency, but it is impossible to clean behind the plates.

In general, box temperatures run 10 to 15 degrees Fahrenheit above the surface temperature of the plates.

Use of Fans to Augment Air Flow

Since air flow is a key ingredient in the box-temperature equation, you can lower the temperature of the box by moving air over the plates with a small muffin fan. These can be thermostatically controlled and have a tremendous impact on box temperature. If you are shy on surface area or want to use a plate with increased eutectic temperatures, a small fan is the way to go.

Plate Location

Fridge plates are best located, when possible, at the top of the box. This keeps them in the warmest spot, where they are least likely to maintain a frost line, and where thermocyling of the air will be most effective over the plate.

This has the further advantage of removing the possibility of freezing vegetables that inadvertently sit against the plate.

Freezer plates are also best mounted on the top, although sides work well, too.

The very best approach, however, is to use the freezer plate in a freestanding installation, where it is a divider in the center of the box. This allows you to get at both sides to clean. It also gets the best air flow around the plate.

Built-In Plates

In many instances it makes sense to use a built-in plate, where the plate is actually a part of the fridge liner. You start out with a stainless-steel liner for the box. This forms the inner wall of the plate. An outer wall is then welded to the inner to form an enclosed tank.

The advantage of this type of approach is that the inside of the fridge is a smooth, easily-cleaned stainless-steel surface, with no hidden areas to collect grunge.

The negative is that only a single side of the plate is in contact with the air, so more surface area is needed.

Plate Plumbing

As the liquid refrigerant is pumped out by the compressor, it flows down the liquid lines (small pipes) toward the box. Just before the plate there is an expansion (TX) valve. This TX valve has a small needle valve in a tiny orifice that is operated by a temperature-sensing bulb.

As the liquid refrigerant hits the needle valve, it is turned into a fine spray of extremely cold gas. As the gas flows down through the plate plumbing, it absorbs heat from the surrounding solution.

The temperature-sensing bulb regulates the flow of gas. The proper setting of this, called the Òsuper heatÓ setting, is critical to efficient operation. Being off just a little bit can reduce efficiency by 50 percent or more.

The design of the plumbing run after the TX valve, and how it absorbs heat, has a major impact on your system's efficiency.

The gas inside the tubing runs at 15 to 20 degrees Fahrenheit below the eutectic point of the evaporator coil at the beginning of the cycle. Eventually, this causes the solution closest to the tubing to freeze. This frozen material then forms an insulating barrier so that the compressor has to make even colder gas (and run at a lower, less efficient temperature) to freeze the rest of the solution.

With a true eutectic, all of the solution theoretically changes state at the same time. In reality, this is somewhat true, although there is always some early change of state (freezing) around the plumbing from about halfway through the cycle.

The more surface area there is, the better heat transfer. At the beginning of the cycle this is based on compressor capacity, but toward the end of the cycle it is based on the capacity of the semi-insulated plumbing pipes to absorb heat.

Coil Design

Coil design varies with compressor capacity. The ideal solution is to use an air-conditioning evaporator coil of appropriate size. This coil has a combination of lots of tubing and interspersed aluminum fins, so that there are large amounts of surface area in contact with the surrounding solution. With this approach the partial freezing of the eutectic has much less impact.

The problem is that these coils use copper tubing and aluminum fins. Put all this in a stainless tank and add some salt water, and you have a big battery. However, this approach will work to improve the efficiency of the noncorrosive noneutectic solutions (such as ethylene glycol and water).

Another approach, used by Glacier Bay, is what they call a spider coil. It uses 120 feet (37 m) of 1/4-inch (6mm) stainless-steel tubing. If this were laid out in a single run, the pressure drop would be too great and there would be a loss in efficiency. Instead, they have the tubing manifolded in a series of parallel runs. This results in good surface area with little pressure drop.

Oil Traps

Your refrigerant travels through the system carrying a mist of oil with it. When the plate is designed, it must take into account this oil and make sure there is enough velocity (speed) of the gas so that oil is carried along back to the compressor.

Plates have specific orientations to make sure the oil is pulled through correctly. You will want to observe this when installing them.

Plate Capacity

If you are using an engine-drive compressor, you will want at least a 24-hour holdover time. Twice this would be preferable. However, there are limits, since plates take up interior-storage area.

With electric-based systems, where you have large battery banks, plate size is typically determined by what is needed to make the electric compressor most efficient, the majority of the holdover coming from the battery capacity, which is then replenished by the engine.

Using Evaporator Coils

If you have a large battery bank, consider a system based on a simple evaporator coil. This is the same as the approach used in home systems. You still have the cold gas running through the pipes, only now there is no eutectic. The plates cool the air directly. The heat transfer, evaporator coil to air, is more efficient than what happens in a holding plate where you first have to cool the solution (with the freezing complications) and then the solution cools the air. But there is no storage capacity with this approach except for what is available in the batteries.

Whose Plates?

There are all sorts of claims about plate efficiency, how they are made, and how they compare to the opposition. Assuming you are starting from scratch, there are several basic choices.

Glacier Bay makes plates with very efficient spider coils that use a salt-water-based true eutectic solution. Another approach is available from Grunert. Bob Williams, of Sea Air Land Technologies was involved in the development of these plates, and Bob says they also have a true eutectic solution. An alternative way to go is with galvanized steel plates made for ice cream trucks by the Dole Company. The Dole plates use salt water. For some reason people don't like the galvanized steel. However, I think they are fine. You can frequently find them used.

Finally, you can make up your own plates. It's not that big a deal, and while you will probably not save any money, you might end up with something just the right size. If this is your approach, I would aim toward plates using the Òsuper juiceÓ we mentioned earlier, with air-conditioning finned evaporators.

Testing an Existing Plate

If you want to find out quite easily if your existing plates have a true eutectic, you will need a recording thermometer, available from most refrigeration-supply houses.

Take the probe and tape it onto the middle of your plate. Use an insulating tape so that it only reads plate temperature, not outside air temperature.

If the plate is eutectic, you will see the temperature drop or rise at a steady rate. When it hits the eutectic temperature, the graph will flatten out until the solution has reached the change-of-state temperature, at which point the temperature will again change. If you don't see this flattening of the temperature range, you do not have a true eutectic solution.