High efficiency, simplicity, competitive capital and operating costs, and extraordinarily low NOx. These are among the benefits promised by the Tophat cycle, which uses water injection before and/or during compression as well as recuperation.


The Tophat cycle, employing wet (quasi-isothermal) compression and therefore allowing the use of a recuperator, is a very efficient and economic means to make maximum use of the Joule cycle. Efficiencies of over 60% (LHV basis) are possible with present day gas turbine technology. It can be applied everywhere that combined cycles are now used. Moreover it offers excellent possibilities for other applications where currently open Joule cycles are used, such as in ships, offshore operations, LNG plants, peak shaving plants, etc. With the Tophat cycle, efficient and competitive power generation is possible for any duty from as low as 500 kW up to 400 MW (although for duties below 50 MW the efficiencies will not be higher than 55% (LHV basis)).

At the same time, on the environmental front, there is no other internal combustion engine which has such low NOx emissions as one using the Tophat cycle.

Surprising enthusiasm

It is somewhat amazing that people are so enthusiastic about the high efficiency of gas turbines that operate in an open Joule (Brayton) cycle (see Figure 1). When used for power generation the efficiency is currently at best about 43%, which is considerably lower than what can be obtained with steam (Rankine) cycles or low speed marine diesel engines. These can now run at efficiencies of 45% and 50% respectively. When considering that the latter cycles can use dirty residual fuels and steam cycles can even use coal and that the gas turbine has to run on distillate fuel or gas this implies that there is considerable scope for improving the efficiency of the Joule cycle.

Of course the gas turbine is by far the most elegant of the above mentioned alternatives for power generation when it comes to compactness and ease of start-up, eg in peak shaving operation. However, to improve its efficiency for power generation the only practical solution until now is to employ the sensible heat in the exhaust gases of the turbine to drive an additional steam (Rankine) cycle. This so-called combined cycle can obtain efficiencies of up to 60%, which is much higher than any other commercial alternative. This solution may be good for efficiency reasons but has the disadvantage that the capital cost increases from US$250 to US$400 per kWe. Further the inclusion of a steam cycle makes it uneconomical to build combined cycles for duties below 50 MWe. Moreover such combined cycles are less suitable for peak shaving than an open gas turbine cycle.

The basic problem is that virtually all gas turbines used for power generation feature adiabatic compressors. For industrial gas turbines and aircraft derivatives having pressure ratios of 15-40 this implies that the outlet temperatures of the compressor and the turbine are so close together that recuperation is not an option (see Figure 2). As a result a lot of energy in an open Joule cycle is lost as waste heat in the exhaust gases leaving the turbine. Further, one should be aware of the fact that the energy for the heating of the air in an adiabatic compressor is transferred via the shaft connecting the turbine with the compressor. In other words power that could be converted into electricity with an efficiency of almost 100% is being used to heat the air. Of course, the use of colder compressed air for the combustion requires more fuel in order to obtain the same turbine inlet temperature. However, this heating is accomplished by burning fuel with an efficiency of 100% rather than the alternative of “electric” heating via the shaft that has an efficiency of only about 40%.

In case of a more isothermal compression the waste heat in the hot exhaust gases from the turbine can be used to heat the air leaving the compressor (see Figure 3). This combination of a more isothermal compression and recuperation is an effective means to substantially enhance the efficiency of a gas turbine based power station. Using waste heat for this purpose is a lot more efficient than – indirectly – using electrical heating.

Recuperation can also be used in the case of adiabatic compression when low pressure ratios, of below 10, are used, as for example applied in the OP16 turbine discussed in reference 1.

Isothermal compression has been proposed in the past, and has sometimes been implemented, but it has always implied the use of intercoolers, causing an additional pressure drop, or application of a so-called humidified air turbine (HAT) cycle, featuring intercoolers as well as humidifiers having an appreciable pressure drop. Moreover such solutions have the disadvantage that the compressor has to be split into several machines, which complicates the basically simple Joule cycle.

The Tophat cycle

Consideration of the above mentioned drawbacks of the Joule cycle per se has brought about the idea of the Tophat cycle (Figure 3) where water is injected before and/or during compression resulting in wet- or quasi-isothermal compression (see references 2-7).

The water is injected in such a way that the compressor does not suffer from a parasitic pressure drop. It can either be introduced in the air entering the turbine or inside the turbine, eg via the stator blades. It is injected in the form of very fine droplets with a mean diameter of about 1-3 micron. Such small droplets can, for example, be made by combining flash evaporation with efficient atomisers as used, for example, in Swirl Flash Technology® (reference 6). Small droplets are required as they:

• evaporate in the milliseconds available in the compressor;

• will not cause erosion problems; and

• will not be centrifuged out in the compressor.

For flashing, the injection water should have a pressure above atmospheric and a temperature that is at least 50 °C above the boiling point corresponding to the pressure of the air into which it is atomised. Moreover the atomisation is promoted by the reduction in surface tension at higher temperatures. The addition of a surface-active agent to the water may also enhance the atomisation. The humidified air leaving the compressor then leaves the compressor at the required pressure and saturated with water. In a recuperator the relatively cold humidified air is subsequently heated with the hot exhaust gases leaving the turbine to a temperature of say 10-100°C below the turbine outlet temperature before being routed to the combustor.

Wet compression

The high heat of evaporation of water – a penalty in a steam cycle (pinch problems) – is turned into an advantage in wet compression. Wet compression, as proposed for the Tophat cycle, is not completely isothermal but quasi-isothermal. In practice it implies that the temperature at which the humidified air leaves the compressor varies from about 100 to 170°C for discharge pressures of 8 to 32 bar respectively and starting with ISO air of 15°C. This is illustrated in Figure 2.

Per unit fuel quasi-isothermal compression requires less energy than adiabatic compression. This advantage increases with the pressure ratio as illustrated in Figure 4.

The greatest advantage of a more isothermal compression is that it now becomes advantageous to have a recuperator in which the sensible heat in the gases leaving the turbine is used to preheat the humidified air leaving the compressor. This heat that in a combined cycle is used to drive an additional steam cycle is now used in the more efficient and less costly Joule cycle.

The recuperator

Because of bad experiences in the past, as well as ignorance, there is a widespread negative attitude regarding the use of recuperators (see reference 4). This perception originates from poor performance with recuperators when combined with gas turbines. However, it is often forgotten that recuperators are highly successful in air preheaters employed in furnaces and in the iron and steel industry, where ceramic recuperators are used for intermittent operation (both in temperature and pressure) up to temperatures of 1500°C.

Furthermore, gas–gas heat exchangers are generally unpopular because they have poor heat transfer characteristics. But, regarding the latter point it is worth recalling that in a typical heat recovery steam generator (HRSG), as applied in every combined cycle station, there is a gas–gas heat exchanger in the hottest part, where steam is superheated up to temperatures exceeding 500°C. In the HRSG the poorest heat transfer, and hence the determining factor in terms of heat exchange area and steel required, is that from the hot atmospheric exhaust gas to the heat exchange surface. Therefore heating humidified air instead of water/steam does not substantially increase the heat exchange area of a recuperator over that of an HRSG. The amount of steel required may even be less because the pressure of the humidified air in the Tophat cycle is 3-4 times lower than that of the water/steam in the combined cycle.

Assuming that the Tophat stations will be started up and shut down as frequently as the combined cycle alternatives, thermal shock should not be too much of a problem, particularly as the metal temperatures and temperature cycles are about the same when the preheat temperature of the humidified air is restricted to the superheat temperature (500 – 550 °C) of the steam.

Moreover, the recuperator has a very smooth temperature profile in the steady state. Because the heat exchange is restricted to the exchange of sensible heat (gas–gas and gas–water) the enthalpy supply and demand lines are almost parallel, as is illustrated in the example in Figure 5. Here the humidified air and the fuel gas are preheated to 450 and 500°C respectively and the water used for evaporation during compression is preheated to 200°C. As can be seen the only pinch problem is caused by the water preheat.

The water cycle

The water of the distillate quality required for injection can be obtained by condensing the water in the exhaust gas. This gas has then to be further cooled after it leaves the recuperator. It is advantageous to use a two-stage direct contact condenser. The first stage condensate, comprising 5-10% of the water present in the exhaust gas, contains virtually all the solids present in the combustion air and the fuel that have acted as condensation nuclei. This water can be used as a purge in order to avoid build-up of solid contaminants in the system. The pure condensate from the second stage can then be recycled and used for humidifying the air.

The use of indirectly cooled condensers is less attractive because of the large amount of inert gases in exhaust gas, resulting in large heat exchange surfaces and hence in costly equipment. Sometimes indirect cooling may be economic though, for example when heat is required for heat and power schemes, such as district heating or seawater distillation.

Because hydrocarbon (and in the future possibly hydrogen) is used as fuel a Tophat based power station is always a net producer of liquid water, which can bring further benefits. The water can be used for irrigation, for example, which is clearly important for applications in desert areas and for the production of desalinated water in, eg, offshore applications. If natural gas is used as a fuel the net production of water is about twice the mass of the fuel.

Tophat cycle efficiency

The efficiency of the Tophat cycle is very dependent on the temperature difference between the hot turbine exhaust gases entering the recuperator and the humidified air leaving the recuperator, the water preheat temperature and the pressure ratio over the turbine. This is illustrated for various cases in the table below.

The data in Table 1 and Figure 6 show that with the maximum Tophat cycle efficiencies that can be reached are about equal to those of combined cycles using adiabatic compressors.

Neither mechanical energy losses nor temperature losses were taken into account. Therefore the calculated efficiencies reported in the last line of the table have been consistently reduced by 3%. The resulting efficiencies should mainly be used for comparison. No extra credit has been given to the Tophat cycle for the following advantages:

• The fact that no special measures are required for generating cooling air for the turbine blades as the cool compressed blast leaving the compressor can be used for this purpose. And

• The fact that no measures have to be taken to control NOx in the exhaust gases by using more complex low NOx burners and/or water injection into the combustion chamber.

The “vacuum trick”

The attraction of the Tophat cycle is that it is a simple cycle. However, in case efficiency is important it is possible to increase the efficiency by typically 1-2 percentage points by raising the pressure ratio over the turbine by a factor of about two by expanding to a vacuum of about 0.5 bara and keeping the pressure ratio over the air compressor constant (see table above, Figure 6 and reference 8). This is therefore the pressure at which the gas then passes the recuperator and the condenser(s). The cold gas leaving the (last) condenser has of course to be compressed in order to release it to the atmosphere (see Figure 7).

The increased efficiency stemming from application of the “vacuum trick” arises from three factors:

• The lower compression ratio for the quasi-isothermal compressor implies that the warmest part can be cut off (see Figure 7) and is in fact replaced by the exhaust compressor which compresses a gas that is about 50°C lower in temperature.

• The fact that the outlet temperature of the quasi-isothermal compressor is about 50°C lower makes that more heat can be recuperated in the recuperator.

• The mass flow of the gas to be compressed by the exhaust gas compressor is lower than the gas that would pass through the warmest part of the quasi-isothermal compressor it replaces. The reason is that the gas in the latter compressor contains about 15-20 mol% water whereas the gas leaving the condenser contains only a few mol% water.

A disadvantage of the vacuum trick is that the capacity of the turbine will be reduced by about 50% compared to the operation where the gas is expanded to essentially atmospheric pressure. This has a negative effect on the capital cost of the plant. Another disadvantage is that an additional compressor is required.

Further optimisation of the Tophat cycle is possible by reducing the temperature difference between the turbine outlet and the recuperator outlet temperatures (see data in table) and by changing the temperature of the water injected during the wet compression.

In all cases the efficiency of the power station will increase with increasing turbine inlet temperatures.

Based on the HHV of the fuel the efficiencies would be about 6 percentage points lower.

Station efficiency and NOx control

In terms of NOx control, the biggest advantage of the Tophat cycle is the fact that the stoichiometric adiabatic flame temperatures (SAFTs) are so low. As is well known lower SAFTs result in lower NOx emissions. In the standard Joule cycle higher station efficiencies are obtained by increasing both the pressure ratio and the turbine inlet temperatures resulting in higher SAFTs (see Figure 8). The SAFTs are higher because the higher pressure ratios result in hotter combustion air because of the adiabatic compression (see Figure 2). Using wet compression by itself or in combination with a recuperator, as applied in the Tophat cycle, generally leads to lower SAFTs for stations with a higher efficiency.

The higher SAFTs for the Tophat case compared with the wet-compression-only case are due to the fact that in the Tophat case the blast is reheated in a recuperator.

In Figure 8 SAFTs are plotted against the compressor discharge pressure for various cases: a Joule cycle; a wet compression case; a Tophat cycle; and a case where only isothermal compression is used. The reason for the low SAFTs and the fact that these decrease at higher pressure ratios for the wet compression and Tophat cases is the progressively lower oxygen content and the higher moisture content of the air at higher compressor discharge pressures (see Figure 9).


The high efficiency of the Tophat cycle, 60% or more, makes it attractive for many applications as well as providing a potential alternative to combined cycle technology. As already noted, examples are peak shaving, marine gas turbines, offshore applications and LNG plants, and heat and power schemes. The fact that Tophat cycles can be applied for duties from 500 kW upwards means that they can even be considered for trucks, locomotives, off road vehicles and mining equipment. This development could also have an impact on fuel cell based power stations as competing against gas turbine based power stations with an efficiency of well over 60% will become difficult.


Table 1. Comparison of efficiencies