Jan-2008

Optimising refinery CO2 emissions

How the inclusion of CO2 LP-based modelling costs can change operations in the short term and impact configurations in the longer term

Mike Stockle, Daniel Carter and Lucy Jones
Foster Wheeler Energy Limited

Article Summary

CO2 emissions continue to present refiners with real challenges, particularly where emissions-trading schemes have been introduced. Modelling CO2 as an operating cost, coupled with investment modelling techniques in the LP, allows refiners to identify the optimum level of CO2 abatement for any given CO2 credit price.

The introduction of CO2 emissions trading, and the additional operating cost this could place on refineries, means refiners now face yet another aspect of their operation they may need to optimise to maximise their refinery margin, and hence maximise profitability. Some of the areas in which a charge for CO2 emissions can drive changes in refinery operation, and even in the actual configuration selected for the refinery, may involve the following:
•    Efficiency improvements
•    Fuel substitution
•    Crude substitution
•    Hydrogen production
•    Residue upgrading
•    Carbon sequestration
•    Modelling CO2 in an LP
•    Case studies.

Emissions-trading schemes
The EU emissions-trading scheme is currently scheduled for three phases, with the total CO2 emissions to be reduced during each phase. These phases are scheduled for:
•    Phase 1: 2005 to 2007
•    Phase 2: 2008 to 2012
•    Phase 3: 2013 to 2017.

Other countries are also looking at bringing in similar schemes. The Australian NETS (National Emissions Trading Scheme) cap-and-trade scheme is due to start in 2010 and is likely to be linked with a similar scheme in New Zealand. Mandatory state-level trading schemes are also under discussion in the US, and some voluntary schemes are already in place.

During EU Phase 2, reductions in CO2 emissions in the UK are to be borne mainly by the large electricity generators; this is likely to be similar across the rest of Europe. For refiners, the situation should be business as usual. However, it is necessary for all refiners to understand the implications of CO2-trading schemes and how they could be impacted by these schemes in the future.

Pressures in other sectors could also have implications on refiners. For example, potential short-term changes in the cost of power may make integration with a combined heat and power plant more attractive. Another potential impact includes the long-term changes to the fuel markets placing a premium on low-carbon fuels such as natural gas, along with corresponding reductions in demand for high-sulphur fuel oil and coke.

Modelling CO2 in the LP
The impact of carbon emissions trading on a refinery is complex. It impacts a number of areas of refinery operation and can drive refinery configuration for new builds and upgrades. To fully understand the impact, CO2 needs to be modelled in the refinery LP. There are three main emission sources that need to be covered in the model:
•    Fuel for process heating, steam raising and power generation
•    Hydrogen production
•    Coke burn-off from the FCC.

CO2 can be easily modelled in both the major software packages used for LP modelling. Emissions can be modelled as a utility, with two levels being used both to allow for free credits and those credits that have to be purchased. Alternatively, surplus credits can be sold.

The key to successfully modelling refinery emissions is having the three key previously identified areas modelled in sufficient detail to accurately predict the emissions produced. For the FCC, this means having a good prediction of coke formation, and hence CO2 emissions. In the case of the hydrogen plant, it means identifying the CO2 produced in the process and ensuring the CO2 from any additional fuel firing is not double counted. The hardest of the three to model are the fuel systems. To accurately model emissions, it is necessary to establish accurate compositions for the fuel gas streams produced in the refinery (and for imported natural gas) and also to establish the carbon content of any liquid fuels used. For many refinery LP models, this will mean improving the accuracy of the model. Many models simplify fuel gas to a single stream that essentially models the energy content of the gas. To allow emissions to be modelled, it is necessary to look at the compositions of the streams in more detail and model the components of the stream.

Where potential schemes for reducing CO2 are being considered, the LP model can be used in an investment-modelling mode to look at the viability of these schemes and how that viability varies with emissions pricing. Where a number of potential schemes have been developed, the LP can be used to identify the best scheme or a combination of schemes to produce a given reduction on CO2, or it can be used to identify the optimum level of reduction at a given CO2 price.

Efficiency improvements
Improved heat integration and better use of the heat required in many process units has already been looked at by the majority of refineries. However, the introduction of emissions charging means there is a new financial incentive to improve the overall efficiency of a refinery. Furthermore, there may be an incentive for improved efficiency at refineries that previously have been less concerned about saving fuel, as they have had access to supplies of cheap natural gas.