Most compressorized systems operate with oil mixed in with the refrigerant. This oil is required for lubrication of the shaft bearing and, in positive displacement compressors like scrolls or screws, it actually provides the seal that is necessary to effect the compression of the refrigerant. The oil is miscible with the refrigerant, and travels with it throughout the system to provide lubrication and compression sealing. That means the oil is everywhere the refrigerant is—Which can lead to several operation problems that need to be addressed in the design and operation of the system.The first issue is oil transport. The oil travels with the refrigerant whenever the velocity of the refrigerant is high enough to carry the oil with it. This means that the system must be designed to ensure adequate velocities are met at all times. The biggest obstacle to refrigerant flow occurs, however, at the expansion device between the condenser and the evaporator. This is by design a big restriction on refrigerant flow—the greater the restriction, the greater the pressure drop across it and therefore the larger the temperature difference between condenser and evaporator. Its function is similar to that of a flow restrictor in a shower head or an orifice plate in a commercial piping system. Therefore, in order to ensure an adequate flow rate for oil transport between the condenser and evaporator, the pressure between these components must be kept above a minimum. If the pressure difference falls below this minimum, oil starts stacking up in the condenser and the machine will trip out on low oil pressure.
While there are methods that can be used to minimize this problem, such as oil pumps and eductors that siphon the accumulated oil out of the condenser, these are usually only provided on larger tonnage centrifugal machines, and even then there is a practical limit to the operating environment the compressor is designed for. In practice, head pressure control is used where low condensing temperatures are expected, such as fan cycling on air-cooled condensers or bypass lines on water cooled compressors. And even then, oil transport problems are a major cause of chiller downtime, especially in cool climates.
These head-pressure control mechanisms ensure a minimum condensing pressure to assure oil transport—which is good, because this ensures trouble-free operation. But it does this at the cost of energy efficiency.
See, a compressor is very similar to a pump. The total mass flow of refrigerant can be thought of as analogous to the flow rate of the pump and the pressure difference between the evaporator and condenser is analogous to the static head of the pump. In order to reduce energy use on a pump, you can either reduce the flow rate through the pump, or you can reduce the static head. In a chiller system you have the same options, but remember that the refrigerant mass flow rate is what determines the cooling capacity of the chiller. If you want to reduce energy use, but still provide the same cooling capacity, your only option is to reduce condenser pressure. One of the great advantages of using the ambient air as a heat sink is that for over 99% of the year, you have access to air temperatures (wet bulb or dry bulb, depending on the heat rejection of the system) below your design condition. This is why part load ratings on chillers are so much better than peak load ratings.
Head pressure controls, however, artificially limit how low this head pressure can go. Just when you are really getting efficient, the system kicks in a mechanism to keep your system from getting any more efficient! Without the need to maintain oil flow, a compressor can take full advantage of the available head pressure relief and operate extremely efficiently.
The second major issue with oil in your system is that it actually inhibits heat transfer, reducing the overall efficiency of the refrigeration system. In fact, a nominal 3.5% charge of oil in a chiller system equates to about an 8% loss of efficiency from this insulating effect (from ASHRAE 601).
And this only gets worse with time. An ARI study found that this insulating effect increases for the first 5-6 years of operation reducing chiller efficiency by about 20% due to oil fouling of the heat transfer surfaces in the chiller.
These numbers assume a constant oil charge—which is a big assumption. In practice, chiller oil charges often exceed the recommended oil charge by very significant amounts—since a common method of correcting a ‘low oil pressure’ alarm is to add more oil—despite the fact that the usual cause of this alarm is the stacking of refrigerant in the condenser, not any loss of oil in the machine. The same ASHRAE study above sampled many machines in the field and found that the average charge of oil was nearly 13% which equates to a total efficiency loss of 21%!
The last major issue with oil in the compressor is the chance of motor burnout. If a refrigerant system experiences a small leak in a low pressure location, moisture and air can enter the refrigerant system. Enough moist air can react with the refrigerant and oil to form hydrochloric and hydrofluoric acids which then travel through the system to the motor windings and eat away the insulation, eventually causing massive arcing and a catastrophic failure of the compressor motor. This acidic residue will be deposited throughout they system, requiring extensive flushing of the chiller shells before the system can safely be brought back on line. This is not only very expensive, but amounts to a very long downtime in what may be a critical building component.
Developed in Australia in the mid-90’s, the Turbocor™ compressor was born out of a desire to avoid the energy penalty and maintenance headaches associated with oil in the refrigerant circuit of traditional compressorized cooling systems.
These centrifugal compressors utilize a unique magnetic bearing to completely avoid the need for oil in the refrigerant system. These systems require very little maintenance and provide excellent efficiency both initially and many years down the road.
Smardt manufacturers water-cooled and air-cooled chillers exclusively utilizing these innovative compressors. These chillers avoid all of the drawbacks of oil, and eliminate much of the cost of ownership that is commonly associated with chiller systems. Since the bearings do not wear, traditional scheduled stop-major chiller teardowns are unnecessary, and all of the downtime and cost associated with oil (around 50% of the maintenance cost on these systems) are avoided. Smardt chillers provide an owner with excellent efficiency, reliability and economy.
5 comments:
I've always thought orifice expansion is a terrible idea in the common refrigeration cycle. You're converting pressure difference to heat in a system that's designed to remove heat. Wouldn't some sort of turbine at the expansion point make more sense? At best you can use the energy from this elsewhere in the process (connect to a pump right before the compressor to reduce compressor energy), and at worst you can just dump the energy to the environment as friction - at least you're not dumping this energy your refrigeration cycle!
I suppose a side benefit to turbine energy recovery would be to provide a physical mechanism for conveying oil across the expansion point. Of course, the oil issue goes away with magnetic bearings.
The refrigeration cycle can be a little confusing, and the role of the expansion valve is one of the more mysterious parts.
However the valve is actually one of the only parts of the system where no heat is added or removed from the system. It is an adiabiatic process, but is also irreversible: there is a significant entropy gain through the valve.
A view of a pressure-enthalpy chart of a refrigeration process demonstrates this. Since enthalpy is the heat content of the refrigerant, you would see an increase or decrease in the enthalpy across the valve. Instead, this portion of the chart is vertical, showing no change through the TXV.
What the valve does do, however, is allow two volumes of gas with exactly the same enthalpy to exist at greatly different temperatures. The pressure difference across the valve allows Boyle's Law to greatly reduce the measured temperature of the less dense volume of gas.
Correction:
The refrigerant on either side of the TXV is liquid, not gas.
Next time, more coffee.
Take a look at the refrigeration cycle plotted in temperature v. specific entropy (here). Notice how the line between 4 and 5 is slanted - this is because it's an adiabatic process (more precisely an isenthalpic process). The reason it isn't straight down is because you're converting energy contained in pressure to entropy. Pressure is converted into heat which boils off some of the fluid (notice you're close to 50% liquid/vapor in this chart). But we don't want extra entropy in our cycle, so I suggest we mechanically remove that energy via a turbine. You'll end up with more liquid to boil off, which means more capacity and less wasted energy.
You know, after just thinking that you misunderstood the process, I now understand what you are getting at.
In fact, it appears that at least one other group is trying to provide exactly this sort of a system.
It seems most of the problems with doing this are technical, but it is sound in principle. This lecture from Yunus Cengel (whom I was able to hear speak at this year's NYC ASHRAE conference) notes that the main reason is one of expense--He also hints at other problems with this approach, I would imagine you would need to be much more confident in your subcooling, etc to avoid flashing through the turbine, for example.
Note however, that this approach is already used in the gas refrigeration cycle, which appears to avoid a phase change.
Whaddayaknow. You think you know everything, and then somebody opens your eyes to a new concept.
Post a Comment