Kalina-Cycle.pdf

An Introduction to the Kalina Cycle

ASME Int ern ati on al

Reprinted From

The American Society of Mechanical Engineers 345 East 47th Street New York. NY 10017

PWR- Vol. 30, Proceedings of the International Joint Power Generation Conference Editors: L Kielasa, and G. E. Weed Book No. H01077 -1996

An Introduction to the Kalina Cycle HENRY A. MLCAK, PE

AB STRACT1

many design options such as reheat, regenerative heating, supercritical pressure, dual pressure, etc. all of which can be applied in a number of different combinations in a particular plant.

This paper is intended as a primer on the Kalina cycle - a novel, efficient power cycle that uses an ammonia-water mixture as the working fluid. The reader needs no more than a basic understanding of conventional water based Rankine cycle power plants to comprehend the basic thermodynamics, principles and arrangements of Kalina cycle power plants presented in this paper.

Each Kalina cycle system in the family of designs has a specific application and is identified by a unique system number. For example - “Kalina Cycle System 5” (KCS5) is particularly applicable to direct (fuel) fired plants. “Kalina Cycle System 6” (KCS6) is applicable to gas turbine based combined cycles and “Kalina Cycle System 11” (KCSI 1) is applicable to low temperature geothermal plants. There are a host of other systems which are applicable for other fuels or heat sources such as municipal waste, waste gas stream in processing plants, solar and even nuclear.

INTRODUCTION Compared to the conventional century-old Rankine cycle, a Kalina cycle power plant may offer efficiency gains of up to 50 percent for low heat energy sources such as geothermal brine at 150°C to 210°C. Gains of up to 20 percent may be realized for higher temperature heat sources such as direct fired boilers and exhaust gases from a gas turbine. i.e. bottoming cycle of a combined cycle plant.

But first, to really benefit from the discussions on the Kalina cycle, an understanding of the Rankine steam cycle fundamentals and basic thermodynamics is necessary.

THE RANKINE CYCLE

It is likely that Kalina cycle plants will even cost less to build than Rankine cycle plants of equal output. Up to 30 percent savings for low heat applications and up to 10 percent savings for direct fired or bottoming cycle plants have been approximated.

In the Rankine cycle, water is heated in a boiler (or a heat recovery steam generator) to make high temperature, high pressure vapour or steam. This steam passes through and turns a turbine as it rapidly expands and cools to a low temperature, low pressure exhaust. The turbine drives a generator thus producing electricity. The steam from the turbine exhaust is then condensed to water using a cooling medium such as ambient air, or water from a pond or cooling tower. The ambient air or pond is referred to as the heat sink. The condensate is then pumped back to the boiler to repeat the process. A schematic of this simple Rankine cycle is shown in Fig. 1.

The Kalina cycle is principally a “modified” Rankine cycle. The transformation starts with an important process change to the Rankine cycle - changing the working fluid in the cycle from a “pure” component (typically water) to a “mixture” of ammonia and water. The modifications that complete the transformation of the cycle from Rankine to Kalina consist of proprietary system designs that specifically exploit the virtues of the ammonia-water working fluid. These special designs, either applied individually or integrated together in a number of different combinations, comprise a family of unique Kalina cycle systems. This is somewhat analogous to the Rankine cycle which, in fact, has

Rankine cycle plants are inherently poor performers. For example, the largest and most modern coal fired Rankine cycle power plants are generally no more than 35 percent efficient. That is, of the total energy in the fuel that is burned, only 35 percent of it is converted to electricity.

1

This paper has been reprinted in SI units in 2004 with permission of the author

1 of 11


Fig. 1 Schematic Rankine Cycle Power Plant The 65 percent energy loss is due to a host of reasons. Approximately 15 percent energy loss is due to moisture in the fuel, heat radiation from the furnace walls, low temperature flue gas from the furnace and auxiliary (parasitic) power. (Kalina cycle plants will have losses similar to these therefore further discussion on these particular losses is not pertinent to the topic of this paper.)

Fig. 2 Second Law Efficiency Limit

In a power cycle, the highest quality energy form available is the heat source and the lowest quality energy form available is the heat sink. Consider a hypothetical ideal power plant with an infinite heat source at 1000°C. The shaded area in Fig. 2A graphically represents the absolute total work from this power plant while operating in a heat sink environment of absolute zero temperature.

The Rankine “cycle losses” account for the remaining 50 percent. This loss is basically the energy that is still in the steam at the turbine exhaust. Although this steam has an enormous amount of energy, it is at such a low pressure and temperature that it is useless. This energy must, therefore, be thrown away to the heat sink just to condense the steam to water so it can be pumped back to the boiler.

Recall the relationship °C + 273 = °K for absolute temperatures. Therefore, absolute zero is 0°K or -273°C.

A quick review of thermodynamics will show why these cycle losses are so large.

Since a heat sink temperature of absolute zero is only possible in deep outer space, all of the work indicated in Fig. 2A cannot be realized here on earth. The shaded area in Fig. 2B indicates the portion of the absolute work that must be forfeited because our heat sink temperature is much higher than absolute zero. Another way to look at Fig. 2B is that amount of absolute total work that our heat sink, acting as a heat source itself, could do in an absolute zero environment. For our example, a heat sink temperature of 10°C, (283 degrees above absolute zero) is used.

FIRST LAW, SECOND LAW & EXERGY The “first law” of thermodynamics, often referred to as the law of conservation of energy, deals with nothing more than the “accounting” of energy. That is, all energy, in whatever form, must be accounted for. For example, in a heat exchanger heat energy loss by one stream is equal to the heat gained by the other. In a power cycle, the heat input to the system is equal to the sum of work and any waste heat produced or discharged during the process.

The difference is shown by the shaded area in Fig. 2C. This shaded area shows the energy available to do work. For an ideal engine, it represents the maximum potential work from a given heat source in a specific heat sink environment.

The above stated 35 percent cycle efficiency is a “first law efficiency” because it is no more than a simple measure of the “work” component performed by the input heat energy. (Power plant “heat rates” are just another way to slate the first law efficiency.)

Notice how the heat sink temperature “limits” an engine from achieving 100 percent conversion of the heat energy to work energy. The ratio (T2 – T1) / (T2- T0) is the “second law efficiency limit.” Using absolute temperatures and T0 = 0, the second law efficiency limit (or our example is:

While the first law tells us “how much,” it doesn’t tell us “why.” Why are the losses what they are? The answer to this question is crucial for one’s understanding of a more efficient cycle.

(1273 – 283) / 1273 or, 77.8%.

This leads us to the “second law” of thermodynamics. This law basically says that “work” will, or can be done by an energy medium as it goes from a high temperature to a low temperature inside a heat engine such as a turbine.

“Potential work” and “exergy” are other terms for this second law efficiency limit. Now suppose the heat source temperature is changed from 1000 to 250°C. The exergy for this heat source is an amazing low: (523 – 283)/ 523 or, 45.9%.

2 of 11


You can visualize this by pencilling in the changes in Figures 2A and 2C. This tells us that, even for a perfect engine, most of the energy in this low temperature heat source would be forfeited to the heat sink. (This should help to explain why it is more difficult to exploit low temperature heat sources for generating electricity where heat sources temperatures are 150°C or lower.) Let’s suppose again that on a hot summer day the heat sink temperature raises from 10°C to 50°C. The exergy decreases from 77.8 percent to 74.6 percent for the 1000°C heat source and decreases from 45.9 percent to 38.2 percent for the 250°C heat source. As you can see, either lowering the heat source temperature or raising the heat sink temperature, lowers the exergy. Conversely, either raising the heat source temperature or lowering the heat sink temperature, raises the exergy. If one could, somehow, directly use the heat source and heat sink in an ideal heat engine, the above stated second law efficiency limits would also be the first law efficiencies for our hypothetical plants. And, of course, these efficiencies would be the highest possible.

Fig. 3

But unfortunately this is not the case. Heat engines like today’s turbines need a medium to go between the heat source and heat sink. This medium is a working fluid.

Second Law Effici ency

For our example, let’s assume that, economically, the working fluid can be condensed at a temperature no lower than 30°C, or 20°C above the heat sink temperature of 10°C.

A turbine in a cl osed loop power cycl e has direct c ontact with only the working fluid. This means that the turbine work is determined, not by the actual heat source and heat sink temperatures, but by the temperatures of the working fluid.

The work that this working fluid is capable of producing in an ideal turbine is shown by the area between the hot working fluid temperature and the cold working fluid temperature. See Figure 3B. Notice that this area, and therefore the work is now less that the potential maximum work shown in Fig. 3A and previously shown in Fig. 2C. “Destruction of exergy” has taken place because of the working fluid.

Go back to our 1000°C infinite heat source and 10°C heat sink example that has a second law limit efficiency of 77.8 percent. Let’s introduce a working fluid and some real life design limitations into our hypothetical plant. To do work, our working fluid needs to be heated by the heat source. But 1000°C is way too hot to handle with current pipe materials. The current metallurgy technology limit in power plants is around 600°C. For the sake of this exercise, let’s assume that the working fluid boils at a constant temperature of 600°C as it absorbs energy from the 1000°C heat source.

The ratio of the “work” produced to the “total heat source energy” is: (873 – 303) / 1273 or, 44.8 %. Now don’t confuse this with a new second law efficiency limit. i.e. exergy. This 44.8 percent is actually the “first law efficiency” of our ideal turbine doing work with a working fluid. This tells us that our hypothetical power cycle is producing much less work than the potential work which was previously calculated to be 77.9 percent. This exercise shows the importance of keeping the working fluid temperature as high as possible as it absorbs heat and as low as possible as it rejects heat.

After producing work in the turbine, our working fluid will have to give up its remaining heat to the heat sink via a heat exchanger. An important aspect of heat exchanger design must be explained at this point. A basic parameter in sizing heat exchangers is called pinch point. This is simply the minimum temperature difference or temperature driving force between fluids. If there is a large temperature difference between fluids, the hotter fluid can easily transfer energy to the cooler fluid with very little surface in a heat exchanger. However, if the temperature difference is small, the surface (and cost) of the heat exchanger increases exponentially. A point is reached where it becomes cost prohibitive to further reduce the temperature difference between the two fluids.

The percent of the potential work converted to actual work (in our example 44.8 / 77.8, or 57.6 percent) is a measure of the “second law efficiency.” To clarify further, 57.6 percent of the potential work in the heat source is actually converted to work. This second law efficiency makes for a good comparison of power cycles because it eliminates measuring something we have little control over - the heat sink.

ANALYZING THE LOSSES Now we can begin to analyse why cycle losses are what they are. In the previous exercise, two types of losses prevented all of the exergy from being converted into work. These are technological and thermo-economic losses. Refer again to Fig. 3 as these are briefly explained below. Technological Losses Our working fluid was limited to only 600°C because we really do not have piping material that can practically handled higher temperatures. Losses in exergy that occur because of limits

3 of 11


like this are called technological losses. Losses due to our imperfect turbines, pumps and compressors are other examples of technological losses.

Now, note that even with these losses, the area of Fig. 4B, and thus the actual work could be larger if the working fluid paralleled the heat source and cooling water curves. However, since the working fluid water boils and condenses at a constant temperature while the heat source and cooling water have variable temperatures this is not possible. These differences account for the structural losses in a power cycle. As previously stated, reducing these structural losses is the goal of Kalina cycle designs.

Advances in technology are still being made but at a much slower pace than they were decades ago because of diminishing returns (more research effort for incrementally fewer gains in efficiency). The Kalina cycle does not solve or reduce these technological losses; however it is important to note that any technological advances made to benefit the Rankine cycle will also benefit the Kalina cycle. Thermo-economic Losses

This finally brings us to the lifeblood of the Kalina cycle - an ammonia-water mixture working fluid.

The cold working fluid was set 20 degrees C above the heat sink temperature because of economics. For our example, we assumed that reducing this 20 degrees difference any further would have increased the incremental capital cost of our heat exchanger above the value of an incremental increase in power output. Losses in exergy that occur because of limits like this in boilers, condensers, cooling towers, etc. are called thermo-economic losses. Thermo-economic decisions will still have to make in Kalina cycle plants in the same basic manner that they are currently made in Rankine cycle plants.

AMMONIA-WA TER MIXTURE Ammonia-water mixtures have many basic features unli ke that of either pure water or pure ammonia. A mixture of the two fluids behaves like a totally new fluid altogether. There are four primary differences. First, an ammonia-water mixture has a varying boiling and condensing temperature. Conversely, both pure water and pure ammonia have constant boiling and condensing temperatures.

Structural Losses The groundwork has now been laid to address the real focus of the Kalina cycle - increasing cycle efficiency by reducing structural losses.

Second, the thermo-physical properties of an ammonia-water mixture can be altered by changing the ammonia concentration. The thermo-physical properties of water and ammonia are fixed.

In the real world, heat sources are not infinite. Power plant heat sources such as hot gases and geothermal brine are sensible and limited. That is, the temperature of the heat source cools as it gives up its energy to the working fluid. On the other hand, working fluids in a Rankine cycle absorb much of this heat at a constant temperature due to the constant boiling nature of a pure component like water.

Third, ammonia-water has thermo-physical properties that causes mixed fluid temperatures to increase or decrease without a change in the heat content. The temperature of water or ammonia do not change without a change in energy. The final difference is not really a change in a basic feature, but rather an important change in a fluid property. This is the freeze temperature. Water freezes at a relatively high temperature of 0°C, while pure ammonia freezes at -78°C. Solutions of ammonia-water have very low freezing temperatures.

On the turbine exhaust side, we have a heat sink environment that is essentially at a constant temperature. However, a finite cooling medium of water or air is used. This finite medium increases in temperature as it absorbs heat from the steam working fluid exiting from the turbine. Conversely, the steam being condensed does so at a constant temperature.

Phase Diagram

Figure 4 shows a more accurate graphical depiction of the Rankine cycle process. The area in Fig. 4A represents the potential work of a given heat source and heat sink, while Fig. 4B represents the actual work of a Rankine cycle. The Rankine cycle actual work is superimposed over the potential work. First note the metal limit and pinch point that reduce the actual work. These are the technological and thermo-economic losses previously discusses.

The essence of the Kalina cycle takes advantage of the first feature change - the ability of an ammonia-water mixture, at any given pressure. to boil or condense at a “variable” temperature.

Fig. 5 Ammoni a-Water Phase Diagram

Fig. 4 Structural Losses

4 of 11


Varying Boiling Temperature

This is in distinct contrast to the “constant” boiling/condensing temperature of a pure fluid like water.

As previously stated, the heat input to a power plant’s working fluid is from sensible/limited heat sources. Therefore the temperature of the heat source drops as it gives up its energy to the working fluid.

Ammonia has a low boiling and condensing temperature compared to water. Therefore, ammonia in a mixture with water is the more volatile component of the two. This means that when an ammonia-water liquid is heated, mostly ammonia will boil off first, i.e. distillation will start to occur. Conversely, when an ammonia-water vapour is cooled, mostly water will condense first.

Because of the variable boiling temperature, the temperature rise of an ammonia-water mixture, in a counter-flow heat exchanger, more closely follows the straight line temperature drop of a sensible heat source.

This unique feature is illustrated in a phase diagram shown in Fig. 5. This diagram plots temperature vs. ammonia-water concentration for ammonia-water mixtures at 550 kPa. (Each phase diagram plot is for a specific pressure).

This is illustrated in Fig. 6, which shows the temperature profile of a liquid ammonia-water working fluid being heated by a stream of exhaust gas. The working fluid, at 3450 kPa - 70 percent ammonia-water liquid, has an initial temperature of 38°C and is heated to a 260°C superheated vapour. The stream of exhaust gas has an initial temperature of 288°C.

Point I at 156°C is the saturation point of pure water. This is the temperature where water boils or steam condenses at a pressure of 550 kPa. Similarly, point 2, at 6.9°C is the saturation point of pure ammonia at 550 kPa.

As the ammonia-water solution at 38°C is heated, it reaches saturated liquid point first at 93°C and then saturated vapour point at 184°C. The vapour is then superheated another 76°C to reach its final temperature of 260°C. The exhaust gas in this example is cooled from 288°C to approximately 93°C as it gives up its energy to the ammonia-water.

Now notice the two curves in between. The bottom curve is the saturated liquid, or boil point for the different concentrations of ammonia-water. This is where initial vaporization occurs when the mixture is heated or where complete condensation occurs when cooled. The top curve indicates the saturated vapour, or dew point where complete vaporisation or initial condensation occurs.

Now using the same available heat source and the same temperature pinch point, the heat absorption characteristic of 3450 kPa pure water will be examined next.

When an ammonia-water mixture is vaporized or condensed, a phase diagram conveys a lot about the process. For example, consider a sub-cooled 70 percent liquid ammonia (in water) mixture, a common mix in the Kalina cycle. When heat is applied, the mixture starts to vaporize at 21°C, shown as point 3 in the diagram. As heat continues to be applied, the temperature rises and more of the mixture (remember mostly ammonia at first) vaporizes.

As shown in Fig. 7, the water is heated from its initial temperature of 38°C to its saturation temperature of 242°C. As heat continues to be absorbed, the saturated water “boils” and goes from saturated liquid to saturated vapour without a change in temperature. The saturated vapour or steam is then superheated to its final temperature of 260°C.

After the solution starts to boil , but before it is full y vapori zed, the solution actually has two separate components - vapour and liquid. For example, at 66°C, the 70 percent average solution is shown at point 4. The vapour component of this solution is shown at point 5. As indicated, this vapour contains a rich solution of 97 percent ammonia (and 3 percent water). The liquid component at point 6 contains a lean 36 percent ammonia in water solution. Using the lever-arm principle, the fraction of the total mass that is 97 percent ammonia rich vapour i s ‘given by the relation: (pt4- pt6) / (pt5-pt6) (70- 36) / (97 - 36) or, 55.7%. The fraction of the total mass that is 36 percent ammonia liquid is given by the relation: (pt5 - pt4) / (pt5 - pt6) (97 - 70) / (97 - 36) or 44.3%. Further heating continues to raise the temperature of the solution. And in so doing, vaporization of the liquid fraction continues until all of the mixture reaches a saturated vapour state at 116°C, point 7. Additional heating results in superheating the ammonia-water vapour. This process is completely reversed when ammonia-water vapour condenses as heat is removed.

5 of 11


Fig. 6 Heat Exchange Process - Heat Source to 3450 kPa Am mo ni a-Water Sol ut io n

Fig. 7 Heat Exchange Process - Heat Source to 3450 kPa Water

Fig. 8 Heat Exchange Process - Heat Source to 690 kPa Water

Fig. 9 Heat Sour ce Process - Heat Sourc e to 55 kPa Water

It is clearly shown in Fig. 7 that the high constant boiling temperature of water limited the amount of energy that could be extracted from the heat source.

Water would have to be lowered to a vacuum of 55 kPa before it could extract the same amount of energy as the 3450 kPa ammonia-water working fluid. See Figure 9. Note how the average temperature of this working fluid is lowered even further by the low constant boiling temperature of the 55 kPa water.

Also note the final temperature of the exhaust gas – almost 260°C. The water cannot utilize the rest of the energy in the flue gas from 260°C on down.

So, although a high average working fluid temperature is desired to increase efficiency in the Rankine cycle, a lower pressure/lower saturation temperature may be necessary to simply increase the amount of heat extracted from a sensible/limited energy source. Now you should be able to see the kind of compromises that are made in Rankine cycle plants.

As you can see, the 3450 kPa water absorbed only 15 percent of the total energy absorbed by the ammonia-water. Now it doesn’t even matter how efficient the rest of the plant can convert this fluid to work because 85 percent of the available energy is not being used. So let’s try to absorb more of the heat source energy by lowering the boiling temperature. Another plot, Fig. 8 shows the same comparison for water at 690 kPa. As can be seen, lowering the pressure (and therefore the saturation temperature) increases the amount of heat that can be extracted. In this case, water at 690 kPa extracts 62 percent of the amount extracted by the 3450 kPa ammonia-water working fluid.

Such compromises are reduced in the Kalina cycle. As illustrated in these comparisons, one can get both relatively high average temperature and high energy transfer with ammonia-water as the working fluid. An interesting side note - designers of larger bottoming (Rankine) cycles in gas turbine based combined cycle plants often increase cycle efficiency by using two or even three different pressure levels to stair step through the heat absorption process. The highest pressure water extracts energy from the highest temperature exhaust gas, while the lower pressure water extracts energy from the lower temperature gas.

But wait. While more of the heat source energy was indeed extracted, lowering the boiling temperature of the working fluid lowered the average temperature of the hot working fluid. And from our previous discussions, we know that lowering the average hot working fluid temperature lowers the second law efficiency.

6 of 11


This increases both the average temperature of the working fluid in the HRSG and the total amount of energy extracted from the gas while still making available a high pressure steam for the turbine. The lower pressure steam is admitted to the turbine at different inter-stage points. An ammonia-water mixture working fluid that parallels the heat source temperature profile eliminates the complexity of these multipressure systems.

This brings us to the next advantage of the Kalina cycle variable condensing temperature makes possible a high degree of energy recuperation. Refer back to the phase diagram in Fig. 5 for a 70/30 mixture at 550 kPa. Point 7 depicts the 116°C saturated ammoniawater vapour at the turbine exhaust while point 3 depicts the 21°C saturated ammonia-water liquid at the condenser outlet. Note that the working fluid must drop 95 degrees C in the total condensation process from turbine outlet to condenser outlet.

Fig. 11 Heat Condensation

2070 kPa

Exchange

Process

-

Recuperation-

550 kPa

Figure 10 shows a schematic of a fundamental Kalina cycle design that capitalizes on this temperature difference. The condensed 21 °C ammonia-water at the condenser outlet is pumped so a recuperative heat exchanger to absorb heat from the 116°C turbine exhaust flow while, at the same time helping to condense the exhaust flow.

Fig. 10 Schematic – Recuperation – Condensation Varying Condensing Temperature Both “steam” in a Rankine cycle and ammonia-water vapour in a Kalina cycle will exhaust from a condensing turbine at or near their respective vapour saturation points. However, given equal condenser cooling water temperatures, the ammoniawater will have a significantly higher pressure and temperature than the steam.

A temperature profile curve of this process is shown in Fig.11. In this example, half of the energy in the turbine exhaust fluid normally loss so the heat sink is recuperated. This figure shows how the variable condensing working fluid of a Kalina cycle makes it possible for a low pressure, high temperature stream to transfer heat so a high pressure, low temperature stream.

The higher pressure is a result of ammonia being more volatile than water. The higher temperature is a result of the variable condensing temperature feature of an ammonia-water mixture.

Now, let’s focus on the final condensation process.

For example, given a condenser supplied with 16°C cooling water, steam in a Rankine cycle should exhaust from a turbine at approximately its saturated vapour conditions of 4 kPa and 29°C. This steam will condense to its saturated liquid conditions at essentially the same pressure and temperature. With the same cooling water, ammonia-water vapour will exit a turbine at its saturated vapour conditions of 550 kPa and 116°C. This vapour, however, will not fully condense until cooled to its 550 kPa saturated liquid temperature of 21°C.

Refer again to Fig. 4 and note how the condensed water temperature depends on the “exit” cooling water temperature, not the “inlet” cooling water temperature. Again, this is due to the constant condensing temperature of steam Therefore, work is loss in a Rankine cycle due so the “structural” difference between the working fluid and the cooling water. Figure 11 shows the heat exchange process for an equivalent condenser in a Kalina cycle. Note how the final ammoniawater working fluid temperature depends on the inlet” cooling water temperature - not the higher “exit” temperature, thus keeping the average heat rejection temperature of the working fluid low.

If you apply the concepts presented in the first part of this paper, you should realize that this high 116°C turbine exhaust temperature in the Kalina cycle would tend to increase the average heat rejection temperature and, as a result, decrease, not increase the cycle efficiency. This would be exactly right if the exhaust fluid went directly to the condenser - but it doesn’t. The high temperature of the ammonia-water makes its energy “available” to do further work while in the Rankine cycle the low temperature of the steam makes its energy “unavailable” for further work.

7 of 11


Fig. 13 Phase Diagram – Absor ption -Condensation

Fig. 12 Schematic – Absorption-Condensation There is also a secondary benefit of this Kalina cycle design. Since the temperature rise of the cooling water for the ammonia- water can be higher than that for condensing steam, the cooling water flow rate for the ammonia-water can be less than that required for the steam. This means smaller cooling towers, pumps and piping, along with lower auxiliary power requirement for pumping and cooling Lower fans. This can result in significant capital cost savings for the cooling system and further increases in the “net” power plant output for the Kalina cycle. Specific Composition Change Fig. 14 Energy-Mass B alance – Heat of Mixing

The Kalina cycle has a design option that can effectively reduce the pressure and temperature of the vapour at the turbine exhaust.

This bit of design genius not only reduces the turbine exhaust pressure, but reduces the exhaust temperature as well. As Fig. 13 shows, the temperature of the ammonia-water vapour at the turbine exhaust is reduced from 116°C, point 3, to 88°C (point 4).

Taking a page out of an ammonia absorption refrigeration design book, the Kalina cycle has the ability to readily change the thermo-physical properties of the working fluid in different parts of the cycle by changing the ratio of the ammonia-water mixture.

Overall Composition Change

A simplified arrangement for this part of the Kalina cycle is depicted in Fig. 12. In this example, a 70 percent ammonia working fluid exits the turbine. This fluid first passes through a recuperator as explained in the previous section. After the recuperator, a very lean mixture of 34 percent ammonia from a distillation system is sprayed in the stream. This spray changes the concentration of the turbine exhaust stream from 70 to 45 percent ammonia. The stream is then condensed to a final temperature of 21°C.

Another feature of the Kalina cycle is that the composition of the ammonia-water mixture, and therefore the thermo-physical properties of the working fluid can be readjusted throughout the entire system if operating parameters change. For any given Kalina cycle arrangement there is an optimum ammonia-water mix for a given cooling (heat sink) temperature and a given heat source temperature. If one or both of these temperatures change, the mix in the cycle can be easily adjusted to again optimize the cycle efficiency for the new conditions.

This absorption-condensation process can be graphically explained by Fig. 13. This figure shows two ammonia-water phase diagrams - one for 550 kPa and one for 205 kPa. As shown, if the fluid was condensed at 70 percent concentration and 21°C, it would have a condensing pressure of 550 kPa (point 1). By changing the concentration to 45 percent, the condenser can - at the same heat sink temperature - condense the turbine exhaust at 205 kPa (point 2).

For most plants, the cooling temperature changes significantly with the season of the year and within a particular season, the temperature will change moderately every hour. While it may not be practical to change the mixture every hour, the mixture can certainly be changed on a seasonal basis. The mix can be changed to optimize plant efficiency for an expected average cooling temperature during a certain period of time. Most fuel fired plants will not see a change in the heat source temperature, however plants utilizing other heat sources may. A plant that utilizes waste heat from a process pl ant may see a temperature change due to a process output change. A geothermal plant may see its heat source temperature degrade with time as the resource area is developed and more

8 of 11


of the geothermal steam or brine is used. Again, as these conditions change over time, the ammonia water mix can be adjusted to give optimum performance.

Just a small amount of ammonia in water greatly reduces the freezing temperature. For example, even a lean 25/75 ammonia-water solution has a very low freezing temperature of -51°C.

The ability to change the mixture, and thus the thermodynamic properties of the working fluid offers an extra degree of control simply not possible in conventional Rankine power cycles. Heat Release of Soluti on Another unique feature of am monia-water properties occurs in mixing. Two separate streams of ammonia-water can be mixed together to form a single stream that has a temperature that is actually higher or lower than either one of the two individual streams.

Fig. 16 A Schematic Ar ran gem ent ; -15°C d b

-

Winter

Condenser

Series

As a result, condensation temperatures of ammonia-water fluids can go down to arctic temperatures. Air cooled condensers do in fact have to be used to exploit this advantage. Kalina cycle plants utilizing water cooled condensers would, of course, be subjected to the same 0°C temperature limitation because of freezing of the cooling water. Fig. 15 Phase Diagram – Heat of Mixi ng

Air cooled condensers do give a lacklustre performance in warm climates. It is a fact that air cooled condensers do not benefit from the evaporative cooling that cooling towers provide. Thus plants with air cooled condensers will have lower efficiencies at summer temperatures. But as with every project, specific site conditions, water availability, seasonal peaking load, etc. must be considered to determine if a particular design is justified.

Figure 14 shows an energy-mass balance for a simple mix. As shown, the two incoming streams are at 15.6°C. The outgoing stream is at 37.8°C. This phenomenon is caused by a reaction referred to as heat of mixing. As the name implies, heat release is a result of mixing and occurs when a lean solution and a rich solution are mixed together.

Kalina cycle plant designers must therefore weigh the winter time advantage against the summer time disadvantage, or they might consider another design option that w not even be considered for Rankine cycle plants. This is the installation of a water cooled condenser an air cooled condenser arranged in series. This arrangement is possible in a Kalina cycle plant again because of the variable condensing temperature.

The phase diagram in Fig. 15 should help to explain this phenomenon. This diagram plots enthalpy vs. ammonia-water concentration for mixtures at 689 kPa. Notice how the mixing line in this example crosses isothermal lines. The advantage of this feature is in heat transfer. For example, if an external heat source or an internal heat recuperation process stream was at 38°C, no heat transfer to the combined stream in Fig. 14 could be possible because there would be no temperature difference between the two streams. However, this same source of heat could transfer heat into each stream individually because the temperature of each is 22 degrees C cooler.

Figure 1 6A depicts this arrangement for winter operation. The turbine exhaust flow, after first going through a recuperator, would next go through a water cooled “first stage” condenser. Here, the exhaust flow would continue its condensation process started in the recuperator. The lowest temperature of the ammonia-water out of this condenser would be 13°C or so. (This is limited by the cooling water approaching freezing conditions in the cooling tower.)

Operation Below Freezing Temperatures We have established that lower heat sink temperatures increase the exergy, i.e. potential work of a heat source. However, water based Rankine cycle plants cannot take advantage of this additional potential when the heat sink temperature drops below 5°C or so. This is, of course, because plant operators will start to control the water temperature to prevent freeze damage. So even though nature, by lowering temperatures below freezing, gives an opportunity to increase cycle efficiency, Rankine cycle plants simply cannot take advantage of it.

The fluid would next flow to the air cooled “final stage” condenser where final condensation would occur. A final temperature of around -3.9°C could be obtained with a dry bulb temperature of -15°C. This is about 16.7°C lower than had only a water cooled condenser been installed.

Kalina cycle plants, however, can take advantage of this opportunity. Ammonia has a very low freezing temperature.

9 of 11


All this points to the fact that the innovation and technology of the Kalina cycle is in the process itself, not in the equipment or materials. There are no major equipment developmental hurdles. Even for binary geothermal plants, where most plants use a hydrocarbon base working fluid, ammonia-water has other major advantages over its competition. First, ammonia-water has higher heat transfer rates than hydrocarbons. Size of heat exchangers, therefore, will be smaller and less costly. Second, hydrocarbon fluids have higher molecular weights than ammonia-water. As a result, ammonia-water can produce equal power outputs at significantly lower flow rates. And finally, the hazards of working with ammonia-water are significantly less than with hydrocarbons. So what about the hazards of working around ammonia? Ammonia is actually a very common, widely used, and regularly handled chemical. There has been decades of working with ammonia in hundreds of synthesis plants, thousands of ammonia absorption refrigeration systems as well as countless fertilizer feed stock distribution centres. Even in conventional power plants, ammonia has been used for decades to control the pH of the working fluid. Ammonia is also being used more frequently for NOx emission control systems and inlet air chillers on gas turbines.

Fig. 16 B Schematic - Summer Condenser Series Ar ran gem ent ; 38°C db , 18°C wb For a 70 percent ammonia-water mixture, the exhaust pressure is effectively lowered from 411 kPa, the saturation pressure of 13°C to 228 kPa, the saturation pressure of -3.9°C.

Large quantities of anhydrous ammonia are handled on a regular basis in farming communities. Farmers and ranchers have many uses for anhydrous ammonia which result in “direct” releases to the environment. As a nitrogenous fertilizer, pure ammonia is injected 10 to 15 cm below the surface of the soil at a rate of 100 to 450 litres per acre. This gaseous ammonia is adsorbed on the surface of clay and humus particles. Anhydrous ammonia is also used to prepare protein feeds for cattle and sheep, as a herbicide to kill pesky weeds and as a defoliant to hasten the shedding of cotton leaves to facilitate mechanical harvest.

Figure 16B depicts the arrangement for summer operation. Through proper valving, the role of the condensers would be reversed. The “first stage” condenser would be air cooled and the “final stage” condenser would be water cooled. A site with a 38°C dry bulb temperature and 18°C wet bulb temperature would produce cooling water at around 23°C. Ammonia-water coming out of the air cooled “first stage” condenser would be approximately 49°C. The final condensing temperature coming out of the water cooled “final stage” condenser would be around 28°C - about 21 degrees C lower than had only an air cooled condenser been - utilized. Here, the exhaust pressure is lowered from 1262 kPa, the saturation pressure of 49°C, to 703 kPa, the saturation pressure of 28°C.

This extensive industrial and agricultural experience has resulted in a maturity of the safety standards applicable for working with anhydrous ammonia. These standards have a proven track record in ensuring safe and environmentally responsible ammonia handling facilities.

MISCELLANEOUS

Due to the distinctive, pungent suffocating odour and irritation properties, even at low concentrations, ammonia is self alarming and actually serves as its own warning agent. No person will voluntarily remain in an area that has injurious concentrations of ammonia. The strong odor ensures operators maintain a “tight” plant.

Conventional axial flow steam turbines can be used in Kalina cycle plants. This is possible because the molecular weight of ammonia and water are similar - 17 for ammonia and 18 for water. Turbines for Kalina cycle plants will also be smaller and therefore less expense than turbines in Rankine cycle plants. This is because Kalina cycle plants can utilize back pressure turbines as condensing turbines.

Anhydrous ammonia is flammable but difficult to ignite. It can only ignite in a very narrow concentration range - 16 to 25 percent by volume in air with an open flame. Ammonia will not support combustion after the ignition source is withdrawn. Ammonia is gaseous at atmospheric pressure. It is lighter than air and. therefore, easy to vent off.

Turbines in Rankine cycle plants exhaust to a condenser under vacuum, whereas Kalina cycle plants exhaust to a condenser under pressure. This creates a big difference in specific volumes and therefore a big difference in the size of the flow area in the exhausting turbine.

Ammonia is an ecologically safe industrial chemical when handled properly. It readily combines with water and carbon dioxide to form relatively harmless compounds. It is even reported to benefit the environment by neutralizing acidic pollutants in the air.

For example, the specific volume of a 70 percent ammoniawater mixture exhausting from a turbine at its dew point of 550 kPa and 116°C is 0.326 cubic meter per kilogram. The specific volume of steam at the same condensing temperature of 21°C but at its saturation pressure of 23 kPa is an amazing 54.2 cubic meters per kilogram.

Ammonia is rendered harmless when diluted sufficiently with water. These lean ammonia-water solutions have numerous industrial and household uses, especially as cleansers. Blowdown solutions too lean to be reused in a Kalina cycle plant can be sold back to the supplier of ammonia for these other markets.

There are no special materials required for ammonia-water mixtures. Carbon steel is quite acceptable for low temperature applications. Alloys already familiar to the power industry are acceptable for high temperature service. The only material that is not acceptable is copper or copper bearing alloys.

10 of 11


is an abundant supply of low temperature geothermal energy sources and waste heat process streams that are not economically feasible to develop using current technologies.

CONCLUSION Most efficiency improvement programs for Rankine power cycles generally target only one or two specific designs at a time. Cycle designs like higher pressures and temperatures, or more reheats hope to gain only a fraction of a percentage point in efficiency. These efficiency gains generally come at a very high cost. Consequently, gains are often not economically justified.

The claims of the Kalina cycle technology have been supported by a 3 Megawatt pilot plant in Canoga Park, California. This plant utilizes 543°C waste heat from a gas fired furnace as its heat source. Several Kalina cycle projects are currently in the design stage. Heat sources for these projects include geothermal, gas turbine exhaust gas (combined cycle), coal fired and municipal waste.

The Kalina cycle, on the other hand, presents a host of new ideas to the power industry at one time. Each idea alone has a potential of several percentage points gain in efficiency. And because of this higher efficiency, the physical size of certain plant equipment will be smaller. Smaller boilers, smaller pollution control systems, smaller fuel handling systems, smaller ash handling systems, and smaller cooling systems will be required for a given power output. This, coupled with the fact that the turbines will also be smaller and standard materials can be employed throughout the plant, means that these efficiency gains may actually come at a lower cost than current Rankine cycle plants.

ACKNOWLEDGMENT The author wishes to acknowledge the substantial assistance of Dr. A. I. Kalina. H. M Leibowitz and Y. Lerner of Exergy. Inc. during the preparation of this article.

BIBLIOGRAPHY Kalina. A. 1.. “Combined Cycle System With Novel Bottoming Cycle.” ASME Paper 84-GT-173

The Kalina cycle does not require technological breakthroughs in equipment design. There is only a lack of experience with an ammonia-water working fluid in the power industry. As such, knockout risks associated with the Kalina cycle are minimized.

Leibowitz. H. M. et al, “Start-up and Operating Experience of the Kalina Cycle Demonstration Plant. Technical Paper TP 9227 presented at ASME Cogen Turbo Congress. Houston. Texas, September 1992.

It is important to realize that the Kalina cycle can take advantage of other technological advances in metallurgy, turbines and combustion/emissions control processes that are sure to be made in the future. Since these types of enhancements do not directly affect the power cycle process, Kalina cycle plants can incorporate these advances when they become available.

Marston. C. H.. “Parametric Analysis of the Kalina Cycle.” Journal of Engineering for Gas Turbines and Power. Vol. 112. pp 107-116. American National Standard for Equipment, Design, and Installation of Ammonia Mechanical Refrigerating Systems. ANSI/lIAR 2-1992.

It is also important to point out that the Kalina cycle’s combined higher efficiency and lower cost advantages should make possible the exploitation of new energy resources. There

Safety Code for Mechanical Refrigeration. ASHRAE 15-1992.

11 of 11

Kalina-Cycle.pdf

Recommend Documents