Hydrogen expands opportunities for power generation in refineries

8 April 2004

Environmental and market pressures are requiring the products of oil refineries to be lighter, lower sulphur and richer in hydrogen. This is resulting in a greatly increased need for hydrogen, which in turn has important ramifications for in-refinery gasification and power generation.

After the success stories of cogeneration in the food and paper related industries, where electricity is the byproduct, the hype is now focussing on poly-generation in large processing industries in general and in refineries in particular. It is often forgotten that the most successful application of cogeneration has for decades been district heating in combination with power generation, where the primary product is electricity. Refineries could become a compromise between these two extremes, with power constituting one of the most valuable bulk products.

From oil well to engine

Before embarking on an investigation of the prospects for cogeneration in refineries it is useful to have a look at energy losses associated with getting oil from the oil well to the internal combustion engines where most of it will be combusted. Some relevant data are given in Table 1.

A few observations can be made about these figures. First, the 10% associated gas is flared (and sometimes even vented, thus increasing the greenhouse gas effect of this loss by a factor of 20-25). This gas is essentially methane and is the most hydrogen-rich of the hydrocarbons leaving the well.

Second, the overall efficiency of large ships and aeroplanes is much higher than that of cars. There are four basic reasons for this: operation is continuous rather than stop/go; lower losses in the refinery; more efficient engines; and – relative to the fuel use during their operating life – the energy requirement for making ships and aeroplanes is negligible.

Third, we note the better figures for automotive diesel engines compared with spark-ignition (Otto cycle) engines – because the production of motor gasoline requires more energy use in the refinery and Otto cycle engines have a lower efficiency.

Flows in the refinery

In Figure 1 a simplified flow scheme for a refinery is shown together with numbers that give the fuel required for the process as a percentage of the chemical combustion energy of the feed to that unit. The units with the highest energy requirements (the catalytic reformer and the catalytic cracker) are those that produce motor gasoline. The reason is that high-octane components are hardly present in crude oils. Hence the molecules have to be rebuilt and that costs energy. For diesel engines the cetane number is important but in general no major modifications in molecule structure are required to comply with the specifications for these fuels. The same holds for the kerosene that is used as jet fuel.

As Figure 1 shows, the first major operation in a refinery consists of distilling the crude at essentially atmospheric pressure. The kerosene and gasoil fractions, with boiling ranges of 150-250°C and 250-350°C, generally only require some refining in the form of hydro-desulphurisation before they can be used as jet fuel and automotive diesel respectively. By submitting the residue of the crude distiller to a vacuum distillation heavier fractions can be distilled off without cracking, at moderate temperatures, below 400°C. The residue from the high vacuum unit can be further cracked to lighter products in thermal cracking units. Again, distillates are obtained that can be further processed into the more desirable lighter products such as motor gasoline, kerosene and automotive gasoil. The residues that cannot be processed economically into lighter products end up in the heavy fuel oil that is used in, for example, marine diesel engines.

Trends in refining

Current trends in the refining business are:

• lighter products;

• lower sulphur products; and

• products richer in hydrogen.

The trend to hydrogen richer products is caused by the more stringent specifications for the aromatic content in motor gasoline and the fact that hydrogen richer products result in cleaner engines and cleaner exhausts. The latter requirement is of course also the reason why product sulphur specifications have become more stringent.

The trend towards lighter products is caused by the tremendous increase in road and air transport. Moreover it can be expected that in the future the heavy fuel oil specifications will also become much more severe. Eventually this may result in the elimination of the residual oil markets and therefore the replacement of marine diesel engines by gas turbines.

Last but not least these product requirements have to be met with more heavy and sulphur rich crudes.

All these trends have one common denominator: they require refineries to greatly increase their hydrogen production. This is where gasification and power generation come in the picture.

Gasification can produce the large amounts of hydrogen required for upgrading in, for example, hydrocrackers using the heaviest residues produced in the refinery as a feedstock (the only other source of hydrogen in a refinery is the catalytic reformer but this is often insufficient).

As the hydrogen production must be ensured at all times it is desirable for the refiner to produce a surplus. This can either be used as refinery fuel or for power generation, as is practised at the Shell Pernis refinery in the Netherlands, for example (pictured p 21).

Refinery processes

Apart from catalytic crackers, which are already well optimised and integrated units, a refinery processing unit typically has the kind of flow scheme shown in Figure 2.

The process temperatures range from 350°C for distilling units to 550°C for catalytic reformers. These temperatures are so high that simple cogeneration, as practised in the food and paper related industries where heat is required at temperatures below 250°C, is not an attractive option as such temperatures are already very valuable for power steam generation.

The feedstock is preheated against the process effluents and the top heat required is supplied by a furnace. The products then preheat the feedstock and are after-cooled in a water or air cooler. This line-up has the disadvantage that the flue gases leave the furnace at a temperature of about 400°C for which some application has to be found. That used to be medium pressure and low pressure steam generation but nowadays this heat is used for preheating the combustion air of the furnace as indicated in Figure 2.

This line-up results in a very energy efficient process as hardly any heat is wasted. But exergy wise the process is not very efficient, as heat at a level of 1100-1300°C is used for the lowly duty of heating feedstocks to temperatures of only 350-550°C. Although substantial improvements are already possible in refineries by making better use of existing equipment and processes1 little has been done about this inherent exergy throw-away in the furnaces supplying the top heat for most of the refinery processes.

Most refinery processes having a flow scheme as shown in Figure 2, such as distillers and hydro-desulphurisers, have a very small net heat effect. Exceptions are catalytic reformers, which are strongly endothermic and severe hydrocrackers that are so exothermic that they only require the furnace for start-up.

By combining the knowledge embodied in Figures 1 and 2 it becomes obvious that it is advantageous to route hot bottom products via a small furnace to the next unit thus minimising the use of feed-product heat exchangers. Hence the bottom product of the crude distiller may be routed to the high vacuum distiller. The bottom product of the vacuum distiller to the thermal cracker and the vacuum flashed bottom product of the latter may be directly used as a gasifier feedstock.

An alternative to thermal cracking may be deasphalting using butane or pentane as a solvent. However, the integration with upstream units is more complex because the deasphalting process operates at a lower temperature than a thermal cracker. For this reason only thermal cracking will be considered here.

Advantages of gasification in a refinery

Gasification in a refinery has a number of advantages:

• It can supply all the hydrogen required for upgrading in a refinery.

• Large amounts of export hydrogen can be made available thus paving the route for hydrogen as a major energy carrier.

• It is an ideal outlet for the heaviest residue of the refinery.

• The synthesis gas that is produced may be used for the production of hydrocarbons by Fischer-Tropsch synthesis or for the production of methanol.

• The gas may be used for the production of power in an IGCC (integrated gasification combined cycle) or other power cycle.

• It opens the way to removing essentially all sulphur in the crude in the refinery. When this occurs in a large number of refineries the sulphur will have to be considered a waste product.

• The gasification feedstock can be thermally unstable as there is no storage foreseen between the thermal cracker and the gasifier. This implies that the thermal cracker may run under more severe conditions thus increasing the yields of more valuable lighter products. When the heavy residues are used in a fuel oil pool such deep thermal cracking is not allowed, as it would negatively affect the fuel stability.

• By producing hydrogen all the carbon in the residue may be converted into a concentrated CO2 stream that only requires compression before sequestering (see panel below).

CO2 sequestering options

Refineries, chemical plants and power stations are the only locations where large amounts of CO2 are emitted in one place, obviating the CO2 gathering problem. In conventional furnaces the CO2 is present at a concentration of about 10% (by vol) in the atmospheric pressure flue gas. Gasification however offers the opportunity to remove the carbon in feedstocks as a concentrated CO2 stream at pressures of 8-40 bar and also produce syngas and hydrogen.

Many studies are being carried out to research the most economically and technologically attractive routes to make CO2 available at pressures of about 100 bar (which is desirable for further transport). The most attractive options are those where the CO2 can be used for the enhanced production of oil or natural gas. One of the most thorough studies covering the whole train, including sequestering, has been carried out by Advanced Resources International, BP and DOE in the so-called CCP (CO2 Capture Project).

Specifically aiming at the reduction of CO2 emissions from refineries is the MIGREYD (Modular IGCC Concepts for In-refinery Energy and Hydrogen Supply) study. As in so many studies existing gas turbines are used, which has a negative effect on the overall efficiencies achievable.

Gasification options in a refinery

Gasification in current refineries

Although both liquid residues and petroleum coke may be used as gasification feedstock liquid residues are preferred as they are easier to handle and can be integrated better. Residues can be directly fed to the gasifier from the flash unit in the thermal cracking process as a 400-450°C liquid whereas the solid petroleum coke first has to be cooled down to a temperature of 100-200°C.

This vacuum flashed residue can be gasified with oxygen in existing gasification processes such as Lurgi's MPG, Shell's SGP or the Texaco Oil gasification process. The primary product in a refinery will in virtually all cases be hydrogen for its own upgrading processes. For reasons of simplicity it will be assumed that any surplus synthesis gas or hydrogen will be used for power generation. The production of export hydrogen will therefore only be briefly considered.

Examples of operating plants where both hydrogen and power are produced using these principles can be found in Shell's Pernis refinery in the Netherlands and also in refineries in Italy and Japan (see Figure 3).2 Once the hydrogen demand in the refinery has been met, clean surplus syngas is routed to the gas turbine for power generation.

Advantageously the heat recovery steam generator (HRSG) of the combined cycle is replaced by a recuperator. This is only feasible when (quasi-) isothermal compression is applied that may be accomplished by injecting fine water droplets into the air entering the compressor of the gas turbine. This so-called Tophat cycle has the advantage that all power is generated via the Joule cycle and no steam cycle is required for steam generated in an HRSG (see Figure 4).3 The water injection water may be obtained by condensing the water out of the flue gases leaving the recuperator. It is recommended that the excess heat available from the recuperator is used to preheat this water and the fuel.


Repowering comprises the use of the oxygen rich flue gases from a gas turbine as preheated air for furnaces (see Figure 5). It can in principle be applied in existing refineries but requires substantial modifications.

Using the sensible heat > 400 °C in the exhaust of gas turbines to replace furnaces

A block flow scheme for this option is given in Figure 6. The sensible heat above 400 °C in the gases leaving the gas turbine is used to replace all furnaces in the refinery.

Comparing the options

Table 2 compares hydrogen and power production for the various options in a non-fuel-oil refinery, with 50 000 t/d crude intake.

These figures make the following assumptions:

• All heating values given are lower heating values (LHV).

• The LHV of the crude is 42 000 MJ/ton.

• The vacuum flashed thermally cracked residue available for gasification embodies 15% of the mass of the crude, corresponding to 13.9% of the total combustion energy in the crude (5850 MJ/ton).

• The hydrogen required for upgrading is 1.5(mass)% of crude. This corresponds to 4.2% of the total combustion energy in the crude (1770 MJ/ton). This is the extra hydrogen required over that supplied by the catalytic reformer.

• The total furnace duty for distillation, hydro-desulphurisation, hydrocracking, etc amounts to 5% of the total combustion energy in the crude (2100 MJ/ton). This is heat at a level above 400°C, as lower level heat for preheating can be recovered using the exhaust gases from furnaces and gas turbines once their temperature has dropped below 400°C.

• The total power consumption in the refinery is 62 kWh per ton of crude (222 MJe).

• Only gas firing is considered. 3% of the heat of combustion of the crude or 1260 MJ/ton crude comes available as byproduct gases.

• Gas turbines have an inlet temperature of 1350 °C and a pressure ratio of 16 where furnace heating takes place with >400°C heat in the exhaust gases of the turbine and of 32 in the case of repowering. Polytropic efficiencies of turbines and compressors are 90%. Combined cycles have efficiencies of 58% (LHV).

• There are no imports of natural gas or power.

In existing refineries the most likely option is to route the residue to a gasifier and produce hydrogen via conventional processes and power by means of a standard IGCC. Using syngas as fuel in a combined cycle offers advantages over the use of hydrogen because in the former case there are no losses in the CO shift. Potential export power for a 50 000 ton crude intake refinery is 511 MWe for hydrogen and 545 MWe for syngas when all 15% residue is gasified. Firing hydrogen is only sensible if the CO2 is sequestered. The latter operation also requires considerable compression energy resulting in an additional penalty and is therefore not a very attractive option to reduce CO2 emissions. Alternatively instead of being used to generate power, the hydrogen may be exported. The potential is about 670 ton/d for the refinery considered.

In case of repowering or when the >400°C sensible heat in the exhaust of a gas turbine is used as a heat source to replace furnaces, no export hydrogen has been considered but only power export.

Repowering of all furnaces in the refinery results in the potential export of power in the range 800-1000 MWe, depending whether a classical combined cycle or Tophat cycle is used. This is much more than in the previous case but requires modifications in the refinery.

Using the sensible heat >400°C in the turbine exhaust for replacing furnaces results in an even larger export power production of 1100 to 1200 MWe (for classical combined cycle or Tophat cycle), corresponding to about 4.5-5.5% of the heating value of the crude intake. It also requires substantial modifications in the refinery and will probably only be considered for new refineries. The additional power production of 200-300 MWe compared with repowering can de explained by the fact that there is less exergy throw-away because no high level heat at 1100-1300°C is being used to heat the oil fractions but only heat in the range 400-650°C.

Alternative fuels

With heavier feedstocks the percentage of vacuum flashed thermally cracked residue will rise above the 15% used in the example given above, which will increase the potential for power production in refineries. Increasing the residue from 15 to 30% on crude will, for a 50 000 ton/day refinery, result in power production in the range 2500-3000 MWe. Such high residue percentages are common with tar sands derived oils.

The same large increase in power production could be obtained in conventional refineries when they import Orimulsion or coal, thus increasing the residue part of the total fossil fuel package and hence the amount of gasifier feedstock. It may not sound attractive to introduce coal but when it is fed to the gasifier as a coal-residue slurry the use of coal is less cumbersome. Texaco and Dow both have successfully proven the use of coal-water slurries as a gasifier feedstock and coal-residue slurries are probably easier to handle and certainly have a higher heating value!

Heat belts

A common problem with many of the above schemes for integrating gas turbines into refineries is that practically each furnace would require its own gas turbine. In newly built refineries some furnaces may use a common gas turbine but this puts an additional constraint on plot planning. A solution is the use of a heat belt but then the question that immediately arises is which medium to use as a heat carrier.

Practical maximum temperatures for hot oil systems are about 350°C and even then rather expensive oils have to be used. As this maximum temperature is on average 100-200°C lower than what is required it may be concluded that none of these oils are suitable for the majority of refinery processes. It is surprising that the use of oil fractions from ethylene cracker residues has got so little attention. These oils have been generated at temperatures of well above 700°C and probably contain certain fractions that can be used at temperatures of well above 350°C. Such oils can be produced at a low cost and hence problems with stability can be solved by having continuous make-up and purging of oil from the system.

An alternative option is the use of liquid sodium. This medium has a bad name because it is associated with nuclear power stations and because it generates hydrogen upon contact with water. The first argument is of course merely emotional when it concerns refineries and as regards the contact with water it is observed that this has of course to be avoided. However, sodium in laboratories is stored under petroleum fractions and hence contact with oils in case of a leak should not be disastrous. Further when the sodium is heated with hot flue gases a leak could well result in a fire, but as sodium oxide is a solid this could be handled. Such a fire is certainly less dangerous than a fire involving light oil fractions! An advantage of liquid sodium is that it gives excellent heat transfer and normal steel pipes can be used.

A third option is to use molten salts. This is not very attractive because of the extreme corrosivity and the relatively high melting points, requiring expensive tracing at all times.The use of oil, sodium or molten salts as a heat carrier would reduce the use of steam as a heat carrier. Currently the situation is such that when you pierce a line in a refinery there is a 50% chance that water or steam is pouring out. Using a heat belt would substantially reduce the use of steam in a refinery.


Table 1. Life cycle energy analysis of transport fuels (relative energy units)
Table 2. Hydrogen and power production in a refinery (assuming 50 000 t/d crude intake)

Linkedin Linkedin   
Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.