In February 1997 BOC led an international consortium, Compania de Nitrogeno de Cantarell SA (CNC), consisting of BOC, Marubeni Corporation, Westcoast Energy, ICA Fluor Daniel, and Linde, to bid for the supply of 40 000 metric tons per day (1200 million standard ft3/d) of nitrogen to Pemex for injection to maintain pressure in the giant offshore Cantarell oil field.

Although the site would be provided by Pemex, there was no infrastructure and the only utility which they would provide was natural gas. The site was in a greenfield jungle location in an environmentally sensitive area near Atasta in Mexico. The challenge to BOC was to develop the most competitive solution within the following restrictions:

• environmentally protected site;

• no infrastructure, other than a main road;

• no power available locally;

• no water available locally (the only permitted water source was from 6 km offshore);

• maximum availability (approximately 100 per cent) was required;

• fuel rate to be less than 110 Btu of natural gas per standard ft3 of nitrogen;

• although a scale-up of 40 per cent was acceptable for the ASU supplied, only proven machinery could be used.

The order was received in October 1997 and in June 2000 the first of the four 300 million standard ft3/d modules, including the nitrogen plant, major compressors, power plant and all utilities, was commissioned ahead of schedule. When the final module was tested, in September 2000, the plant output more than doubled the world’s existing nitrogen generating capacity and consumed more than 400 MW of electricity and steam.

Enhanced oil recovery

Enhanced oil recovery is a technique in which gas or water is used to either maintain or increase pressure within the reservoir or to pressurise the oil out of the reservoir. In either case the result is to increase the quantity of oil recovered from the reservoir.

The Cantarell field, which is located offshore, in the Gulf of Campeche, is the largest oilfield in Mexico and the sixth largest in the world. Oil production started in 1979 but by 1996 the reservoir pressure had dropped by 60 per cent to 1520 psia (104.8 bara). Pemex therefore concluded that if reservoir pressure was not stabilised, the economic life of the field would be limited with a consequent loss of oil revenues.

Pemex further concluded that pressure maintenance was required and examined two options: gas injection; and water injection.

Water injection was discarded because, if used, the geological nature of the reservoir could lead to the watering out of some producing wells.

Gas injection was selected as it was considered more effective than water injection in preventing water movement into the reservoir by taking advantage of the gravity drainage recovery process in this location.

Among the major gases considered were natural gas (dry, sweet, mostly methane), carbon dioxide, flue gas (exhaust from power plants, gas turbines, gas engines or heaters), nitrogen, and associated gas (wet, sour, produced gas).

Based on the geology, economics and efficiency of the different gases, Pemex chose nitrogen.

Identifying optimal solutions

In determining the optimum solution, a combination of ASU (air separation unit) cycle design for nitrogen production matched to compressor design, and both matched to the energy cycle, needed to be considered. The objective: to balance the top of the energy cycle – the air compressors, ASU and other infrastructure – and the bottom of the energy cycle – the nitrogen compressor – with the energy available from a simple cycle or combined cycle solution or a cogeneration solution in the most economic manner, taking into account quality, schedule and cost.

Pemex designated natural gas as the fuel of choice so natural gas fuelled gas turbines would provide the basis of the solution. But the use of gas turbines introduces its own restrictions, in that each turbine has a fixed design performance. In addition, this performance varies considerably with temperature and degrades at different rates over time for different types of turbine. The fixed design performance, reduced to accommodate degradation, defines the number of turbines. However, because of the required temperature range, turbines which would be acceptable at the lower end of the range, but would not provide the necessary output at the upper end of the temperature range, were eliminated.

Finally, many of the existing large gas turbines have not been used to drive compressors, which is an issue because of: the potential requirement for a gear; start-up issues; control issues; and rotordynamic considerations.

The above considerations led to the need to study all available large gas turbines and multiples of medium and small gas turbines and how they best fitted multiple compressor and ASU arrangements.

BOC led this energy and machinery optimisation study, which evaluated a number of ASU and compressor arrangements integrated with a number of potential gas turbine arrangements.

To meet the objective of the study, the group first developed a number of generic schemes involving various compressor types with various driver arrangements as well as a number of different simple cycle and combined cycle gas turbine types and arrangements before analysing them technically and commercially against the Pemex requirements.

Analysing the potential solutions

The challenge of developing and analysing a large number of complex solutions required a combination of experience in air separation design, compressor design and gas turbine combined cycle design supported by state of the art simulators.

As the energy providers (gas turbines) were varied by size and by manufacturer, so the power and steam available to the compressors and utilities varied, changing the design of the ASU.

ASU and compressor optimisation

As only 20 per cent of the consumed energy was associated with the separation process itself and the remainder used to drive the compressors and auxiliaries, the philosophy BOC adopted was that this was essentially a machinery application problem and consequently the project would be won by close investigation of the technical margins associated with the major pieces of machinery. In essence, we would lead the compressor manufacturers to the solution we determined to be optimum both commercially and technically within the total project boundaries instead of the individual equipment boundaries.

The keys to providing a solution were people and programs. The task force we selected was a small focused group of experts in compressors, gas turbine combined cycles and ASU process design. Together they had experience totalling about 150 years.

With respect to programs, BOC had been developing and purchasing a number of tools during previous years to enable us to carry out energy optimisation studies and had won integrated energy cycle projects with their use (see “Integrated heat and power plant supplies copper facility at Gresik”, Modern Power Systems, April 2000).

The first tool was the proprietary BOC compressor simulator and database, known as COMPRESS. This is based on compressor stage test stand performance collected and normalised over a period of years. That tool enabled BOC to configure the compressor with its intercoolers and mechanical losses and predict the power expected to be obtained on a test stand for the configured compressor.

It also enabled the investigation of the impact of interstage cooling, cooler approach temperature and interstage pressure drops on power for the myriad of configurations.

Finally, it enabled the investigation of machine frame size limits and speeds on driver configuration permitting both direct drive and geared drive solutions to be investigated.

Using this program the ASU and the main air compressor as well as the nitrogen compressor could be optimised and perhaps most importantly, as it is based on actual test stand data, the risk associated with all the possibilities could be quantified.

The data from this program, combined with prices from BOC’s compressor market price prediction program known as MAPGRAPH, were fed forward into a financial analysis.

Analysing the energy cycle

Using the compressor and ASU simulators we were able to accurately predict the individual power demands for each module and the total demands for all modules.

Then using the gas turbine simulator and financial program we were able to evaluate the following:

• simple cycle aeroderivative and industrial gas turbines drivers;

• aeroderivative and industrial gas turbine cogen cycles;

• aeroderivative and industrial gas turbine combined cycles;

• all steam configuration.

This was done initially for 1500 million ft3/d and then for 1200 million ft3/d of nitrogen resulting in more than 70 such energy cycle analyses, each with their own systems of utilities, which were included in the evaluation.

Table 1 summarises one of the analyses.

Equipment experience

The issue of experience with respect to the various configurations was also addressed. However, this was investigated last because it was important not to eliminate a combination of equipment which may have provided the optimum solution simply because one element was not sufficiently experienced. BOC’s philosophy is that in today’s highly competitive market it is important to understand all possible solutions, their risks and their impact on the financial objectives before a final selection is made.

Solving the water challenge

Because of the site location, use of any existing water was prohibited. The only permitted source was seawater from an offshore location approximately 6 km away. Seawater, of course, could not be used in any direct cooler configuration typically used in ASU design or as boiler feedwater for the power plant. A source of non saline water also needed to be located to satisfy these needs. As BOC’s compressor simulator calculates the amount of water knocked out at each compression stage, we were able to conclude that this was sufficient to provide makeup water to the ASU and boiler feedwater plant, thus eliminating the need for a desalination facility. This innovation was key to being able to use the gas turbine cogeneration arrangement without excessive cost.

The optimum selection

We concluded from our study that the best solution, meeting all the financial requirements including the availability and reliability objectives as well as the client’s experience requirements was the four-train option, as shown in the process schematic diagram (see previous page).

The plant is optimised for the lowest energy at the top and bottom of the cycle. The natural gas is burnt in the gas turbines to create electricity, which, after deducting a sufficient amount to accommodate gas turbine degradation, optimally matches the needs of the main air compressor and the plant auxiliaries. The waste heat from the gas turbines is used in duct fired heat recovery steam generators to provide steam to match the demand of the steam turbines driving the nitrogen compressors and the demand of the ASU pre purification unit (PPU) regeneration heaters. The ASU cycle pressures were varied to permit optimum gas turbine and steam turbine matching.

Important in this solution is the fact that the power demand can be predicted accurately by BOC’s simulators. Consequently, oversizing the gas turbine, as would normally be done in power plant design, is not required.

For the plant’s non-saline requirements, water is extracted from the air through the main air compressor intercoolers, stored and treated then used to provide makeup to the ASU and boiler feedwater circuit.

The main cooling medium is seawater. The gravity-fed makeup water system uses seawater taken from 5 km offshore to a sump on the beach 6 km from the main facility where it is chlorinated. From the beach it is pumped to a 26 cell induced draft seawater cooling tower system. The seawater is then fed through underground pipes to a central sump where it is chemically treated. From there, seawater is sent to the main compressor intercoolers and to the steam turbine condensers, via 10 foot diameter pipes, before being returned to the cooling towers. Blowdown occurs every 1.5 cycles and the water is returned to the same location from which it was taken in the Gulf of Mexico. The temperature rise across the facility is limited to 2 degrees.

The integrated power, steam and industrial gas system consists of the following major components to ensure the availability and reliability requirements can be met:

• Four GE 7EA gas turbine generators each site rated at 75 000 kW with steam injection capability for NOx control and power augmentation, three operating and one on standby.

• Four ABB HRSGs each with supplementary firing capability, three operating and one on standby.

• Four GHH DK 125 extraction condensing steam turbines driving four GHH nitrogen compressors, each rated at 54 000 kW. All operate during normal conditions.

• Four GHH AR120 axial radial main air compressors driven by ABB two pole electric motors each rated at 52000 kW. All operate during normal conditions.

• Four ASUs, each capable of producing more than 300 million ft3/d of nitrogen. All operate during normal conditions.

With the arrangement adopted (see process schematic), if either a gas turbine or a heat recovery steam generator is out of service the nitrogen facility and pipeline/platform will still receive 100 per cent of desired capacity.

The plant is designed for “black start” and operates in island mode, unconnected to any external power grid.

Electric power is produced at 13.8 kV, 3 phase, 0.85 power factor and 60 Hz. It is then transformed up to 68 kV and distributed to various points in the facility. At various substations the electrical power is transformed back down to 13.8 kV for the main compressor drivers and large pumps and then to 520 V for the balance of plant.

Achieving success

The machinery energy cycle developed by BOC was key to winning the project. The overall facility performance tests held in December 2000 showed that the energy consumption per module and overall was about 1 per cent lower than guaranteed and more than 30 per cent lower than Pemex’s expectations as defined in their bid documents.

The recovery of water from the air through the intercooler condensate system and then treating it prior to reuse as boiler make up water has been particularly successful. There has been no requirement to import fresh water since the end of commissioning.

The impact on production from the Cantarell field has been enormous. In a recent interview (see Offshore magazine, September 2000) Pemex provided the following information:

• Due to pressure maintenance, Cantarell has the capacity to produce 1 896 000 b/d, up from 1 560 000 and this will jump to over 2 million b/d by year end 2002.

• 878 billion cubic feet per day of associated natural gas will become available for power production or as feedstock for petrochemicals.

• As of July 2000 the total accumulated investment for the Cantarell on and off shore upgrading project was 4.32 billion dollars and the total estimated accumulated income, due in part to the nitrogen facility coming on stream, had already paid for the investment.

• Bigger benefits are expected in future years when income outpaces future investments.

Impact on EOR, IGCC, and GTL

The successful completion and operation of the Cantarell facility has helped demonstrate the viability of large scale commercial air gas supply schemes with integrated energy cycles.

Until the start up of the Cantarell nitrogen plant, enhanced oil recovery for large reservoirs had essentially been carried out by natural or associated gas reinjection, water injection or CO2 injection. However, Bechtel and Pemex have recently provided figures (see Table 2) which demonstrate, for the Cantarell case, that nitrogen is by far the least expensive solution for enhanced oil recovery. Consequently, other companies are currently developing similar projects.

In addition, the successful completion and operation of the Cantarell facility is an important step forward in the commercialisation of large ASU plant based technologies such as:

• IGCC (integrated gasification combined cycle) power plant technology, which is the cleanest of the clean coal and waste to energy technologies, having virtually no sulphur dioxides or nitrous oxides in the emissions. Such plants use between 1500 ton/d and 10000 ton/d of high pressure oxygen and three times as much nitrogen.

• GTL (gas to liquids), a method of converting inaccessible natural gas assets or associated gas from oil fields, which is normally flared, into saleable products. Such plants are currently being considered and could require up to 20 000 ton/d of oxygen and an integrated energy cycle.

The significance can best be demonstrated by considering the performance of some of the individual blocks in the Cantarell facility:

• The amount of air processed is equivalent to that used in a 14 000 ton/d O2 plant which would be used for a 2200 MW IGCC unit – 2.0 million Nm3/h (1.24 million scfm) of air at medium pressure.

• The amount of nitrogen supplied is equivalent to that required for full nitrogen integration by more than four FA technology gas turbines – 1.34 million Nm3/h (1200 million scfd). Nitrogen pressure at Cantarell is 120 bar but the first compressor body discharges at just over 21 bara which is ideal for gas turbine integration.

• The design and operation of ASUs of a commercial size applicable to IGCC and GTL has been demonstrated. The process column diameter of the ASU for IGCC service would be physically smaller than that of Cantarell.

• Large scale steam integration with the ASU is demonstrated. 800 t/h of HP/MP/LP steam is supplied to the condensing/extraction steam turbine drivers used to drive the nitrogen compressors and also to the process.

• Steam integration with ASU main air compressors was demonstrated at the BOC Gresik plant (MPS, April 2000) where a condensing steam turbine is used to drive the main air compressor. The steam is provided by the gas turbine HRSGs and waste steam from the client upgraded on the BOC facility to turbine requirements.

• Finally, large scale combined seawater cooling and freshwater cooling of the ASU and power generation equipment associated with a large scale IGCC plant is also demonstrated by the Cantarell project.

Table 1 Summary of cycle selection
Table 2. Comparison of gas injection prices for Cantarell