Mar-2011

Hydrogen management in a GHG-constrained refinery

Recently the US Environmental Protection Agency announced that it would regulate greenhouse gas (GHG) emissions from power plants and oil refineries starting in 2011.

Scott Sayles, Bill Fairleigh, Rick Manner, Joris Mertens, KBC Advanced Technologies

Article Summary

This announcement increases the attention on the largest CO2 producers in the refinery and how to control their CO2 production. Refiners have a number of options for reducing CO2 emissions, as needed, once a full accounting of existing emissions has taken place.

This article discusses the various options for reducing CO2 emissions, including concepts such as improved heater efficiency, energy conservation and utility system optimisation. The primary focus is on how a refinery can optimise hydrogen production to reduce its carbon footprint. One of the largest producers of CO2 is the steam methane reformer (SMR). Alternative configurations such as partial oxidation flow schemes with CO2 capture are also considered.

Refinery configuration and CO2 emissions
Existing refineries have various levels of complexities and operating objectives. The simplest refinery type is the hydroskimming refinery. These refineries typically do not need an SMR hydrogen plant, as the hydrogen demand can be adequately met by the naphtha reformer. Margins are often low at hydroskimming refineries unless there is a large marketing advantage, usually based on location. It is usually advantageous to maximise the production of higher-value transportation fuels instead of lower-value fuel oil products. As a result, most profitable refineries are either moderately complex or highly complex refineries that have cracking and conversion units. Hydrogen requirements can vary greatly from one configuration to the next. The configurations considered when generating examples of refinery CO2 emissions are shown in Figures 1 to 5.

A refinery effectively takes low hydrogen to carbon (H/C) crudes and converts them into high H/C ratio products. Two routes are used to accomplish this:
• Hydrogen addition with a gasoil or resid hydrocracker or hydrotreatring
• Carbon rejection primarily with a gasoil or resid FCCU or resid conversion via coking or visbreaking.

Petro-SIM, KBC’s proprietary process simulator, was used to evaluate a number of refinery configurations, including hydrogen addition and carbon rejection refineries. For all of the configurations evaluated, the crude rate and slate were kept the same. The bars in Figure 6 show the percentage distribution of calculated CO2 emissions for the various process areas with each configuration. Below each bar are the calculated tons of CO2 produced per kiloton of crude processed and per kiloton of light products (gasoline plus middle distillates). All of these cases take into account refinery fuel gas produced, and the marginal fuel is natural gas.

A few interesting observations can be made about the relationship between CO2 emissions and the relative extent by which light products are made via carbon rejection vs hydrogen addition. The last case shown (resid HCU) is an extreme example of a hydrogen addition complex refinery. This configuration makes the most transportation fuels as products, but also has the highest CO2 emissions. The fourth case (high conv HCU) is a moderate hydrogen addition configuration, while the third (high cov FCC) is a carbon rejection scenario.

The main difference between the third and fourth cases is that in the carbon rejection case most of the gasoil-range material is processed in an FCCU, while in the moderate hydrogen addition case most of the gasoil-range material is processed in a hydrocracker. Not as much hydrogen appears to be required in the FCCU case, although much of the FCCU products are hydrogen deficient and require hydrotreating. While H2 plant CO2 emissions are less for the FCCU, the burning of the FCC coke is the highest CO2 emission contributor for the third carbon rejection case and, as a result, the overall CO2emissions for both of these cases is about the same, whether one looks at emissions per ton of crude processed or emissions per ton of gasoline plus distillate fuels.

Although the CO2 emissions for the third case are shown to be roughly equal to those of the fourth case, it should be pointed out that this third case assumes a highly efficient first quartile FCCU. Since most FCCUs are less efficient, a complex refinery relying on carbon rejection with an FCCU may emit more CO2 emissions than a moderate H2 addition refinery with an HCU.

Comparing the high conversion FCC scenario (Case 3) with the residual hydrocracker scenario (Case 5) shows a noticeable difference in CO2 emissions between these two configurations. In Case 5, more of the heavier molecules are being saturated with hydrogen to produce lighter products instead of coke. Since the coke is not burned on the site, the refinery’s CO2 emissions are lower for this case. Emissions for Case 3 would be higher than Case 5 if the coke is burned onsite. Thus, the amount of hydrogen is much greater for Case 5 than any of the other cases. Cases 3, 4 and 5 have the most opportunity from reducing GHG emissions by improving H2 plant operation.

Refinery carbon management programme
With the increasing incentive to reduce the emission of GHGs, refineries and petrochemical plants face big challenges in their operation. Typically, refineries produce at least 30% more CO2 compared to best practice. A project in a North American refinery demonstrated that up to 15% emissions could be reduced with projects having a simple payback of three years or less. For regulatory compliance, and before embarking on any projects to reduce CO2 emissons, a carbon management programme needs to be put in place.

Establishing the refinery carbon footprint
The first step for any refinery is to determine the baseline CO2 emissions. In most refinery units, such as an FCCU, it is not easy to measure emissions directly. CO2, which is generated in flares, incinerators and biotreatment plants, is typically not measured. To account for direct measurement inaccuracies and to benchmark against best practices, it is usually worthwhile to use reconciliation and modelling software to check and validate the measurements. For example, simulation models for refinery fuel, steam and hydrogen balances can be integrated into a program like KBC’s CarbonManager to calculate current emissions and estimate future emissions. Employing such tools allows for the establishment of a base line for future CO2 emission reductions.