Oil refineries are one of the most significant sources of pollution falling under IPPC. The Environment Agency pollution inventory shows they are among the largest sources of sulphur oxides, volatile organic compounds and benzene (ENDS Report 368, pp 10-11 ). Tighter air quality standards and the previous integrated pollution control regime have driven substantial emission reductions by the industry - but significant challenges remain.
To assess the scale of the IPPC challenge, ENDS has reviewed two permit applications - BP’s Coryton refinery in Essex, recently sold to Petroplus, and Ineos’s Grangemouth works in Scotland - and compared them with standards in official guidance.
The findings point to a significant gap between performance at refineries and best available techniques (BAT) as defined under UK guidance. Companies are particularly exposed on SOX emissions.
The applications also suggest reluctance at BP and Ineos to invest in BAT. Assessments of applications from other mature sectors such as steel, cement and chemicals have pointed to similar problems (ENDS Report 321, pp 23-25 ).
John Henderson, head of the Agency’s strategic permitting group for the sector, says the quality of applications is variable. Some sites proposed upgrades, but others offered very little.
ENDS invited BP and Ineos to comment, but both declined because their permit applications have yet to be determined.
Confusion over confidentiality
Oil firms submitted their permit applications last August, but the documents have only recently become publicly available. Concerns that they might contain information useful to terrorists, such as fuel tank locations, caused the delay.
Last summer, the Agency refused ENDS’ request for copies of refinery applications, pending a decision by the Environment Department (DEFRA) and the security services over what parts of the applications to keep secret. DEFRA eventually ruled that only site maps should be withheld.
Not all operators seem concerned about the terrorist threat. No request for confidentiality on the grounds of national security was made by the two Welsh refineries owned by Total and Texaco. And the Scottish Environment Protection Agency (SEPA) was able to send a copy of Ineos’s Grangemouth permit application, containing detailed site maps, without delay.
Other companies, such as BP at Coryton, only requested site maps be kept off the public register.
In contrast Esso, part of ExxonMobil, requested its whole Fawley refinery application be kept secret, as did Conoco
Phillips for its Humber refinery. DEFRA has refused, but ConocoPhillips’s application is still not available.
The confusion over confidentiality means members of the public, perhaps concerned about the impacts of a nearby refinery on their health, have been unable to scrutinise the applications or submit comments to the Agency.
Mr Henderson admits the situation is "not best practice", but insists that the public will have the opportunity to comment. The pollution prevention and control regulations require the Agency to give the public 28 days to respond to a permit application.
‘Gold-plating’ emission limits
The oil refining industry’s main concern is that IPPC should be implemented in a way that does not damage the competitiveness. The UK Petroleum Industry Association has raised far-reaching concerns about the economic challenges facing the industry, such as its ability to meet changes in consumer demand and in the quality of crude oil supplies (see box ).
UKPIA accuses the Agency of ‘gold-plating’ BAT emissions levels that were agreed by the European Commission and published in BREF guidance for refineries. The association highlights standards for SO2, NOX and particulate matter from catalytic crackers.
The BREF says BAT is a combination of feedstock hydrotreatment, use of a catalyst additive, and regenerative flue gas scrubbing technology. SO2 concentrations in the range of 10-350 milligrams per cubic metre are achievable with these techniques.
The Agency’s guidance, which is based on the BREF, endorses the same techniques, but gives a tougher range of 50-100mg/m3.
On particulate matter, both documents agree that BAT includes the use of multistage cyclones followed by an electrostatic precipitator or scrubber. The BREF says 10-50mg/m3 is achievable, while the Agency sets the benchmark at 10mg/m3.
For NOX, the BREF gives a range of 40-150mg/m3 achievable through design modification, feedstock hydrotreatment and the use of selective or non-selective catalytic reduction. The Agency initially proposed a much tougher level of 10-20mg/m3, but relented with a final requirement more in keeping with the BREF.
Ian McPherson, UKPIA environment director, says the Agency is being "over zealous" in deriving indicative BAT values and appears to be going beyond the range of achievable emissions levels recommended in the BREF. He argues it is unreasonable of the Agency to expect existing plant, which may be many years old, to meet standards achievable only by new plant.
In response, Mr Henderson says the values in the Agency’s guidance have been shown to be achievable at European and US plants. Companies will not be forced to meet the benchmarks, but must justify why indicative BAT is not achievable at a particular site and suggest alternatives that improve performance.
The differences between the BREF and Agency guidance are perhaps explained by the variation in performance of EU refineries. A performance survey at 43 EU refineries in 2004 found significant differences in emissions levels. Performance was best in Germany and Scandinavia, but poorer in France and Spain. The UK was seen as "average" (ENDS Report 361, p 8 ).
This variation may explain the wide range in emissions levels in the BREF. The process to produce the BREF was riven with disagreement between member states over BAT for the sector (ENDS Report 322, pp 47-48 ).
The Agency is also trying to help the sector by developing a more flexible approach to regulation through site-wide emission ‘bubbles’ for SOX. The approach is already used in the combustion sector (ENDS Report 368, p 25 ).
Refineries have many different release points such as stacks and vents. Rather than setting a limit on each one, an overall bubble limit could be set, giving operators more freedom over where to make improvements to achieve it. Specific emissions limits on the most significant sources may need to be retained.
One problem is that most refineries, including BP Coryton and Esso Fawley, have not provided sufficient data on the contribution of each source of SOX to enable the Agency to set a bubble limit. The Agency has requested further information using a statutory ‘schedule 4’ notice.
SOX from BP Coryton
Meeting IPPC standards for SOX will be a colossal challenge for refinery operators. The main refinery sources are catalytic cracker and sulphur recovery units. Boiler plant may also be a key source where refinery oil is used as a fuel, in addition to refinery gas.
At BP Coryton, the main stack from the catalytic cracker emits SOX at concentrations of 2,680mg/m3. The annual mass release of SOX from this stack is more than 6,500 tonnes. There is no end-of-pipe abatement for SOX.
Remarkably, SOX concentrations from the sulphur recovery unit stack are 20,460mg/m3. The mass release is 2,300 tonnes per year. The units do not have tail gas treatment, recommended as BAT in BREF guidance, which can reduce concentrations to 400-2,000mg/m3.
BP’s air quality modelling predicts Coryton’s emissions are exceeding two air-quality objectives, including the 15-minute mean.
BP’s application acknowledges it is not BAT for SOX emissions. The company has provided a detailed assessment of the options to reduce the pollutant. BP’s position is that use of a SOX reduction additive and improvements to the sulphur recovery unit’s efficiency is BAT for the site. The company proposes a trial to establish the effectiveness of additives, which it says is still uncertain.
This preferred combination of techniques would reduce SO2 in ambient air to 246-322 micrograms per cubic metre, compared with the 15-minute mean air quality standard of 266µg/m3. The annualised cost is estimated at £880,000 to £2 million.
In contrast, fitting a wet gas scrubber to the cracker and sulphur recovery unit would achieve a concentration of 96µg/m3.
BP acknowledges that a scrubber is the best environmental solution, but balks at the estimated £4 million annual cost.
"The UK market for refined products is oversupplied, resulting in reduced margins," says BP. It would be "difficult to deliver the required financial performance if excessive abatement demands were required through the IPPC permitting process. This could limit the continued viability of the site."
Crucially, much of the expenditure on scrubbers is a capital cost, while SOX reduction additive is largely an operating cost.
BP says: "Capital-intensive abatement techniques are extremely difficult to justify as economically viable in the context of the industry sector relative to abatement techniques that increase operating costs, and could threaten the long-term viability of the installation."
However, BP’s plans to sell Coryton could also explain the company’s reluctance to invest in the site. BP announced the sale last June as it was preparing to submit the permit application. The deal was completed in early February.
BP’s argument is striking given the recent report into the Texas City disaster in the US and the incidents at Grangemouth in 2000. The report highlighted a cost-cutting culture, among other safety failures, as a cause (ENDS Report 384, pp 6-7 ).
In any case, Agency guidance makes it clear that departures from BAT may not be justified on the grounds of individual company profitability.
A flaw in BP’s assessment is that the company has compared the cost of abatement with benefit of achieving air quality standards. The Agency’s BAT assessment guidance requires costs to be compared with the tonnage of pollutant abated to give a more objective comparison. The Agency has required BP to submit this information through a schedule 4 notice.
Ineos also admits it does not meet BAT standards for SOX emissions, but makes a poor job of justifying what improvements are needed. A significant shortcoming is Ineos’s failure to clearly show whether its emissions meet indicative BAT standards. In the BREF and Agency guidance, BAT is expressed as a concentration limit, for most substances in mg/m3.
Although Ineos has carried out an impact assessment using the Agency’s recommended H1 methodology, it fails to compare pollutant concentrations with indicative BAT levels.
In contrast, BP provides detailed tables comparing concentration levels of each release with indicative BAT levels.
Another flaw in Ineos’s application is the absence of a BAT options assessment, again in contrast to BP. Ineos proposes an improvement programme aimed at "closing the gap" between current performance and BAT, but this is simply a list of action points.
To improve cracker performance, Ineos proposes trialling a SOX-reduction additive in 2007. It will also consider the feasibility of fitting a flue gas scrubber. Any upgrade would not take place until 2012.
Improvements at the refinery have helped air quality. In 2003, there were 167 exceedances of the 15-minute mean, falling to five in 2005. This mirrors Grangemouth’s reduction in mass SOX releases from 6,200 tonnes in 2003 to 3,500 tonnes in 2005.
Nevertheless, dispersion modelling shows the site has potential to exceed air quality standards by a considerable margin, especially if Ineos switches to more sulphurous crude. It predicts SO2 concentrations in ambient air could reach 1,270µg/m3 compared with the 266µg/m3 15-minute mean and 166µg/m3 compared with the 24-hour mean of 125µg/m3.
NOX and dust
BP and Ineos are less likely to be troubled by IPPC’s requirements on nitrogen oxides and particulate matter (PM10). Nevertheless, the performance of some processes may need improvement.
At Coryton, the cracker only achieves a concentration of 140mg/m3 of PM10, despite using multistage cyclones. The Agency’s guidance says electrostatic precipitators are BAT on crackers and can reduce emissions to 10mg/m3. Despite this, BP’s BAT assessment concludes that no improvements are needed.
BP says it meets BAT for NOX. The main source is a combustion process which is fitted with steam injection that reduces peak concentrations of NOX from 420 to 150mg/m3. The abatement was non-operational until a recent overhaul.
Nevertheless, BP says further action is likely to be needed to reduce mass releases of NOX as a result of the EU large combustion plant Directive.
At Grangemouth, Ineos identifies several improvements to reduce NOX emissions, but it is difficult to see how it arrived at its conclusions.
An issue of public concern at Grangemouth is noise. In its application, Ineos notes there were 92 complains from local people.
The main cause is flaring of emissions from the ethylene crackers, carried out for safety reasons when the plant starts up and shuts down. Increasing plant reliability has helped the company reduce complaints from a peak of 250 in 2001.
In September 2004, the company fitted a ground flare to reduce the intensity of elevated flaring. But it has been taken out of commission and is currently not operational. Ineos does not propose reinstalling the plant or making other improvements to reduce flare noise. A response to the application from a local community group presses SEPA to apply BAT to minimise disturbance.
In sum, the permit applications from BP and Ineos are disappointing given corporate commitments - from Ineos for instance - "to meet or exceed all relevant legislative requirements and minimise the effect of our operations on the environment."
The pressure is now on the Agency and SEPA to insist the operators justify their performance and commit to comprehensive improvement programmes where they are found lacking.