Cool changes

Refrigeration is not something that can just be tacked onto a boat. It is a major power consumer, particularly in warm climates where DC refrigeration (the most common) can often account for more than 50 percent of a boat’s total DC demand. A refrigeration unit has to be carefully integrated into a boat’s available power supplies, which, in turn, are going to be a function of how the boat is used. Weekend voyaging, for example, will mean shore power is available at the slip all week, whereas a boat used for long-term offshore cruising will not.
A boat with a 24-hour-a-day AC generator can obviously use household-style AC refrigeration units; less common is holding-plate AC refrigeration. In the absence of a constant source of AC power, there are three popular refrigeration choices: engine-driven (with holding plates), constant-cycling DC, and holding-plate DC.
AC refrigeration
Constant-cycling: AC refrigeration requires a constant source of AC power. Given a 24-hour-a-day generating situation, it is the cheapest and simplest option, using household appliances. This situation, however, should not be confused with the AC power now increasingly made available by a DC-to-AC inverter. Any attempt to run AC refrigeration from an inverter will result in substantially greater energy losses than in most other systems — it would be better to install a DC system.
Holding-plate: AC refrigeration requires intermittent operation of a generator. It employs one or more tanks filled with a solution that has a freezing point below that of water. The tanks are fitted to the icebox. When the refrigeration unit is running, it freezes the solution in the tanks, and then the unit shuts down. The tanks slowly thaw out, absorbing heat as it leaks into the icebox, keeping the box cold. When the tanks have almost defrosted, the unit is turned back on to refreeze them. In a well-designed system, even in the tropics the tanks will hold down the temperature in the icebox for 24 hours, and will then require no more than one to two hours refrigeration running time to be refrozen.
However, running an engine in order to spin a generator that is used to power an electric motor that spins a refrigeration compressor that could have been driven directly off the engine in the first place involves a lot of unnecessary energy losses! What is more, the directly engine-driven compressor will, in most instances, have a higher output than the AC unit, and therefore a potentially reduced holding plate pull-down time (depending on whether or not the holding plates can absorb additional compressor output — see ‘engine-driven refrigeration’ below).
In other words, it is rarely worthwhile to run a generator simply for refrigeration purposes, but if the generator is to be run for other loads regularly enough and long enough to support the refrigeration system, and has surplus capacity that can usefully be diverted to AC refrigeration, then holding-plate AC refrigeration is attractive.
Engine-driven refrigeration
Holding plates are used once again, but this time the refrigeration compressor is directly belt driven from an engine instead of being driven by an electric motor, eliminating the intermediary energy losses (see, for example, Sea Frost,
On the surface of things, the beauty of such a system is that given the horsepower available with most engines, a large and powerful refrigeration compressor can be used, minimizing the holding plate freeze-down time and engine-running time. Given that the compressor is directly driven by the engine, engine-driven refrigeration would also appear to be the most efficient.
Holding plates, the limiting factor: Unfortunately, for refrigeration systems of the size commonly found on pleasure boats, neither of these propositions — power and efficiency — is necessarily true! The reason is related to the rate at which heat can be withdrawn from a holding plate — i.e., the rate at which the plate can be frozen. This is largely a function of the surface area of the evaporator tubing in the holding plate, the spacing of these tubes (the closer together, the faster the rate of heat removal), and the temperature differential between the refrigerant in the evaporator coil and the solution in the holding plate.
Except in large systems with multiple holding plates, the rate of heat removal is always well below the nominal refrigerating capability of the compressor on the system: in other words, no matter what size refrigeration compressor you put on the engine, only a certain amount of its capacity can be used, and the holding plate freeze time cannot be accelerated beyond a certain point. This, in turn, means that regardless of compressor capacity, for most systems the engine will have to be run at least an hour a day, and almost always longer than this (in the tropics, two hours or more is common).
If a high-output alternator is put on the same engine and wired to a large enough battery bank to absorb the alternator’s output, even with the added inefficiencies in DC refrigeration (the alternator charges batteries which supply an electric motor that spins a compressor which could have been turned by the engine in the first place), sufficient energy can be put into a battery bank to meet the daily refrigeration load in less engine-running time than is needed for engine-driven refrigeration: in other words, if minimizing engine-running time is a goal, the DC approach often does a better job of capturing the available power from the engine than does the engine-driven approach. The net result is that, although from a refrigeration perspective, DC refrigeration is less efficient. From a whole-boat energy analysis, it will be more efficient on those boats where the goal is to minimize engine-running time.
Consider also the fact that the typical cost of buying even a small marine diesel engine and having it professionally installed is anywhere from $12,000 on up, with a life expectancy of anywhere from 5,000 hours of running time on up. As a result, the capital cost — excluding maintenance and fuel bills — of running the engine is almost always at least a $1 an hour, and all too often can be $3 an hour or more. That’s a significant overhead if the engine is being run solely for refrigeration purposes — minimizing engine-running hours can result in substantial long-term savings.
As noted, this argument breaks down if the engine will be regularly run for other reasons (as it is, for example, on many charter boats), in which case engine-driven refrigeration can be tacked on as just another load. The argument also does not hold in the case of large iceboxes with multiple holding plates that have the capacity to absorb a high compressor output. In this case, engine-driven refrigeration may be the best choice. However, it does suffer from one other major drawback, which is that even at dockside the engine must be run in order to refrigerate.
Sometimes an engine-driven system is combined with a DC or AC system (for use when the engine is not running, or at dockside) by adding a second evaporator coil to the holding plates. The trade-off here is the volume of the holding plate occupied by the second evaporator coil (reducing the plate’s overall capacity, and sometimes reducing the length, and therefore the rate of heat removal, of the primary, engine-driven, evaporator coil), and the added complexity and cost, versus the ability to run the unit at dockside without cranking the engine.

Although expensive, an engine-driven unit will compare very favorably in terms of cost with high capacity DC systems, and even small DC systems if the boat’s batteries and charging systems have to be upgraded to support DC refrigeration.

DC refrigeration

A small refrigeration unit is permanently connected to the ship’s batteries and controlled by a thermostat in the icebox. Every time the temperature rises beyond a set point, the unit kicks on until the icebox is cooled down, and then the unit switches off — it constantly cycles on and off (just as a household refrigerator does).
Aside from household appliances, constant-cycling DC refrigeration is the cheapest initial option. Norcold (, Frigoboat (, Waeco ( and others build drop-in units, similar to household refrigeration units (many of which can also be run off AC power when available), or else a refrigeration unit is added to a purpose-built icebox. Either way, most units currently available use Danfoss BD35, BD50 and BD 80 variable-speed compressors with HFC-134a refrigerant.
These systems have a limited refrigerating capability. Conservatively speaking, I would consider the upper limits to be 7,500 BTUs per day in a refrigerator, and 3,500 BTUs per day in a freezer, although there are plenty of functioning systems that exceed these numbers (up to 10,000 BTUs for refrigeration, and 5,000 BTUs for freezers). This refrigerator number is much higher than I previously recommended, during the time when I recommended against using constant-cycling DC freezers. This change reflects a considerable increase in the capacity of the Danfoss compressors over the past 10 years, coupled with an equivalent improvement in efficiency.
Sea Frost has a constant-cycling unit, the BDxpx A/W, that uses a Danfoss compressor operating on a commercial refrigerant known as 404a. This has a substantially higher capacity than the 134a units, and operates to lower temperatures. As such, it makes an excellent choice for high-capacity refrigerators and/or low-temperature freezers.
Whatever the load, unless there is some kind of a continuous battery-charging device that keeps up with the demands of the refrigeration unit, this kind of a battery drain will cycle the ship’s batteries on a daily basis.
The batteries on a boat used for weekend voyaging can be recharged at dockside during the week, but on those boats that engage in more extended voyaging, if there is no constant battery-charging source (either from running the boat’s main engine on a powerboat, or from a battery charger powered by an AC generator), a powerful DC electrical system is needed to support the refrigeration load. The DC system will include such things as deep-cycle batteries, a high-output alternator, a multistep regulator, and maybe a wind generator with backup solar panels. If not already installed, such a DC system will cost several times more than the constant-cycling DC refrigeration unit!
Constant-cycling DC refrigeration is therefore an appropriate choice on a boat with relatively modest refrigeration needs, one that is to be used for weekend cruising, one that has continuously-operating battery charging devices, or one that already has a powerful DC system. Given a powerful enough DC system, large refrigeration needs can be met through multiple units and iceboxes, or through the use of a Sea Frost BDxpx A/W. If the heat load of a large icebox exceeds the capacity of a single constant-cycling refrigeration unit, two or more units can be installed.
Holding-plate DC refrigeration
Holding plates for refrigeration systems with relatively light loads can be pulled down over time by the small Danfoss compressors found in constant-cycling units (see, for example, Technautics, www.technau; and some from Isotherm,, with some potential efficiency gains.
However, the addition of holding plates increases cost, weight, and complexity (the holding plates require expansion valves, which must be tuned to the system, as opposed to the capillary tubes used in other small DC systems, which require no user interaction), and the plates take up otherwise useable space in the icebox. Holding plates do not maintain as consistent a temperature in the icebox as evaporator plates. The benefits of holding plates over an evaporator plate in small DC systems are arguable (Technautics is an enthusiastic proponent of this approach — see their Web site for a different perspective.)
To take full advantage of a holding-plate system, a larger compressor is needed, commonly driven by a 0.5-hp DC motor (see, for example, Sea Frost, A suitable bank of good-quality deep-cycle batteries can sustain a refrigeration load of up to 0.5 hp for extended periods (one to two hours at a time). The current draw of the unit will be around 40 amps at 12 volts (half this at 24 volts) on a fully defrosted holding plate, tapering down to as little as 20 amps at the end of the holding plate freeze cycle. Such a system will handle substantial refrigeration and freezer loads (up to ten times the small, constant-cycling units). However, it will only function if backed with a continuous DC charging source or a high-capacity DC system (Sea Frost recommends a minimum battery bank of 660 Ah at 12 volts). The refrigeration control circuitry should include the ability to turn the unit on and top it off whenever the engine is cranked. This will optimize engine-running time and minimize the load on the batteries.
A large-capacity DC holding-plate system is expensive, even ignoring the cost of upgrading the DC system should that be necessary, but it has a considerably greater refrigerating capability than a constant-cycling system. It may also be more efficient than the constant-cycling system, or a small DC compressor coupled to holding plates, which means that although it consumes far more power when running, it will provide an equivalent refrigeration capability for less overall power consumption, or else a greater capability for the same overall power consumption. If the DC system includes a large wind generator and/or a large enough array of solar panels, the needs of substantial iceboxes can be met when cruising without having to crank the engine.
Constant-cycling DC
refrigeration vs. holding plate

For many years I was an advocate of high-capacity, holding-plate DC refrigeration for offshore voyaging boats rather than constant cycling. This is because the holding-plate system has had considerably greater capacity and probably efficiency. I have recommended such a system to numerous people, designed one for a number of boats, and put it on my own boats. However, in recent years technology and legal changes have altered the cost/benefit analysis on which this recommendation has been based. In particular:
The relative efficiency of the small compressors used in constant-cycling units versus the larger versions used in holding-plate systems has improved.
The efficiency of the small compressors can be further boosted by a variable speed controller such as the Smart System Controller (SSC) from Frigoboat, or the Adaptive Energy Optimizer (AEO) from Danfoss. (Note that these efficiency gains are made by running compressors for longer hours at lower loads and thus can only be realized on air-cooled units, or those with a through-hull condenser or keel cooler — i.e., without a water pump — because if a water pump is run for longer hours its added power drain cancels out the efficiency gains at the compressor.)
The refrigeration capability of the constant-cycling compressors has increased by more than 50 percent in recent years (and with the introduction of the Danfoss BD 80 went up another 50 percent; Sea Frost’s BDxpx A/W unit is even higher). In the past, in warm climates the available constant-cycling refrigeration units have often been marginal in terms of their capacity to handle even mid-sized refrigerator and freezer iceboxes, and reliably maintain the necessary temperatures for a dependable freezer. This is no longer the case.
Then we have the changes in the insulation world. Improved insulation around an icebox considerably reduces the refrigeration load, enabling this load to be met by a lower-capacity refrigeration unit (see, in particular, Glacier Bay,; and Refrigeration Parts Solution, For those on a limited budget wanting to improve the efficiency of a refrigeration unit, the same increase in efficiency can often be achieved at less cost by upgrading the insulation than it can by investing in an expensive high-capacity holding-plate system.
When you put these things together, and look at them given that the high-capacity holding-plate DC refrigeration systems are commonly four times as expensive as a constant-cycling unit, for many boats it is difficult to justify the added cost. However, it must be recognized that: 1. If vacuum-based super insulation is part of the equation, it can rapidly eat up a large part of the cost savings; and 2. although they are better than they used to be, the constant-cycling units still have a limited ability to handle refrigeration loads — the upper limit for tropical voyaging on (a well insulated) icebox is somewhere around 15 cubic feet for a refrigerator and 5 cubic feet for a freezer.
Larger iceboxes will either need some kind of a holding-plate system, or more than one constant-cycling unit in the icebox. Another approach is to break up the refrigeration load into multiple smaller iceboxes using individual constant-cycling units.
If this were the end of the story, it would make a pretty strong argument for constant-cycling DC units in many applications where previously holding-plate refrigeration was recommended. But the simple lines I have drawn are blurred by hybrid systems such as Glacier Bay’s Micro HPS (Hybrid Plate System).
The Micro HPS is a high-capacity holding-plate system that uses a powerful hermetic DC compressor. This eliminates the maintenance issues associated with externally driven compressors.
The Micro HPS system comes in a compact, pre-charged, skid-mounted configuration that simply needs the refrigerant and cooling lines to be connected — in other words, the installation is no more complicated than that for many constant-cycling units (although it still needs to be vacuumed down, which substantially drives up the installation cost as compared to those constant-cycling units which do not need vacuuming).
It uses downsized holding plates that are somewhere between an evaporator plate and a traditional holding plate, thus reducing the volume and weight of the holding plates. The net result is something midway between a constant-cycling unit and a traditional holding-plate unit, which is reported to have a higher efficiency than a constant-cycling unit (I have not seen any test results to quantify this), a considerably higher refrigerating capability (so long as the holding plates have the capacity to absorb this capability), and less potential for leaks than a traditional holding-plate system using an externally-driven compressor.
If the emphasis is on capacity and efficiency, the refrigeration load may best be met with a hybrid system, although it will take a few years to assess the cost-effectiveness of this approach, and its precise position in the marketplace.
Making choices
Where refrigeration loads are relatively light, and in situations where a boat is used primarily on weekends and plugged into shoreside power during the week (enabling batteries that are discharged over the weekend to be recharged, and the DC refrigeration system to be run off the battery charger at dockside), constant-cycling DC refrigeration is by far the most economical, trouble-free, and easy to install for most boats.
If greater capacity is required in a situation where: 1. engines are run regularly; 2. the DC system is somewhat weak; 3. refrigeration is not required at dockside; then engine-driven refrigeration makes sense. These are the conditions found on many charter boats.
If an AC generator is constantly running, household constant-cycling equipment will be the most economical. If an AC generator is intermittently, but regularly operated, AC holding-plate refrigeration may make sense.
In most other situations, some variant of DC refrigeration is the best choice because it can: 1. minimize engine-running hours; 2. be operated from shore power via a battery charger (whereas an engine-driven system requires the engine to be run at dockside); and 3. be left to run at anchor until the batteries go dead. (One of the design parameters I set myself is a battery and refrigeration balance that will allow the boat to be left unattended for up to a week without the fridge and freezer melting down.) In contrast, engine-driven refrigeration requires someone to be on the boat to run the refrigeration unit every day.

With respect to the DC options, the balance of the argument has definitely shifted in favor of constant-cycling DC refrigeration, even on a hard-core voyaging boat (although the jury is still out on the hybrid systems). Better yet is to have two (or more) units — separate fridge and freezer installations — to provide built-in redundancy. There will still be cost savings over a high-capacity holding-plate DC system. Some of the savings can be put into improved insulation. The end result will be a virtually maintenance-free refrigeration system that is compact, quiet, and reliable; that optimizes the icebox volume and keeps down weight; that has built-in redundancy; and all at a cost that is less than that of a high-capacity holding-plate system.      

By Ocean Navigator