Life-cycle energy efficiency model influences upstream project design

June 22, 2009
Through analysis of life-cycle costs, upstream facilities engineers can evaluate design alternatives such as turbine vs. motor driver selection and establish best practices for various types of facilities.

Through analysis of life-cycle costs, upstream facilities engineers can evaluate design alternatives such as turbine vs. motor driver selection and establish best practices for various types of facilities.

The upstream oil and gas industry historically has focused primarily on equipment availability as a primary design input, which differs significantly from typical designs or operating strategies found in sectors operated on smaller margins, such as refining and chemicals.

The analysis method discussed can benefit preappraise, appraise, and select project stages of a typical offshore development. The primary technique covered involves profiling various options with life-cycle analysis to determine potential energy efficiency and carbon footprint improvement.

This article discusses how these input factors affect decision making early in project planning and execution. A hypothetical analysis of a West Africa floating production, storage, offloading (FPSO) project illustrates the model.

Availability to efficiency

As defined in this analysis, availability is a measure of the time a piece of equipment or system works compared to the required time.1 The requirement for a high overall availability often results in installation of spare equipment, even though the peak rate period may be short compared with the life of the facility.

In this article, efficiency, termed intensity when expressed as a percentage, is the amount of energy expended compared with the energy produced by the facility. Projects have not included efficiency as a primary design metric for several reasons, despite the potential for savings and environmental benefits. Reasons include for not including efficiency include:

  • Lack of a valuation method for fuel gas in developments without access to gas markets.
  • Difficulty in prediction modeling for the life of the field to quantify benefits.
  • Variability in emissions tax regimes and incentives for energy efficiency.

This is not to assert that efficiency is more important than uptime. At current oil prices, uptime return can be about double any savings due directly to lower energy use.2

Some methods discussed for improving efficiency, however, also may increase or at least maintain benchmark availability values.

While it may appear difficult to justify design or operating strategy changes in an environment of relatively high oil prices, one should balance the status quo against the realization that the easiest and cheapest incremental barrel of oil to produce is typically a theoretical one from an existing facility. This production often results from a noncapital expense project.

It should then be possible to quantify and realize some of these operational lessons during the design phase.

This approach differs from others published in that energy efficiency is a primary design metric rather than as an operating facility assessment exercise.

To better understand the issue, we first describe energy use in a broader context across these focus industries, identify a particular area of improvement opportunity and determine the changes needed to effect that improvement.

Energy usage

Upstream and midstream oil and gas are the most energy intensive industries in the world in terms of annual heat input (Fig. 1).3 4 Based on 2005 production of 254 × 1015 btu from oil and gas combined,5 the oil and gas industries collectively consume nearly 20% of the fuel value of produced fluids in compression and pumping, heating, and other processes.

Sectors in the industry use this energy differently (Fig. 2).2 3 Upstream on average expends the majority of energy in fluid transportation.

This varies depending on the type of operation in question. A heavy oil extraction facility in northern climes will have a greater proportion of energy dedicated to heating than an equatorial sweet-oil production facility. In midstream, transportation comprises nearly the entire energy budget. For downstream, more than 90% of energy used is for heating through a direct furnace or boiler application, for steam or other heat medium generation, and for process cooling requirements.

Reduction of required work or losses in the transport segment provides the largest target for energy efficiency improvements in upstream, although heat integration may also have significant opportunities.

The analysis refers to heat integration in the downstream sense, where a design plans and matches process heating and cooling loads to reduce outside requirements and recovers excess process and utility heat for power generation (cogeneration cycle).

Design guidelines

Projects can apply several categories of potential efficiency improvements during the design phase. Generally, one can group these into strategies that reduce the amount of work required, reduce the concomitant losses in performing the work, and reduce fluid losses outside of energy requirements.

Note that some approaches will require multidiscipline support such as optimization between reservoir requirements and facility design. Methods for reducing the energy required for the three strategies include:

  1. Reducing work through:
    • Variation in arrival pressure to analyze horsepower requirements.
    • Smoothing of peak flow to reduce design case requirements (plateau extension).
    • Waste heat recovery (whether for additional power generation or for process needs).
    • Heat integration, such as use of cross exchangers.
    • Use of drag reduction agents to reduce pressure losses in supply or export lines.
    • Offshore deep suction indirect cooling (suction from seawater in the thermocline).
  2. Reducing lost work through:
    • Increased number of pumping-compression units required for peak rates (more flexibility as rates decline).
    • Variable speed electric motors.
    • Variable inlet guide vanes for compressors and turbines.
    • Variable speed mechanical couplings, such as Voith Vorecon.
    • Modern electric motors.
    • Centralized electric generation from larger, higher efficiency colocated turbine generators.
    • Provision of power from outside high efficiency source, such as local grid driven by a combined-cycle gas turbine (CCGT) plant.
  3. Reducing nonenergy-related fluid loss through:
    • Elimination of continuous vent or flare sources.
    • Fugitive emission surveys.
    • Recovery and reuse of spilled hydrocarbons.

Some of the proposed alternatives will have a higher capital expenditure (capex) than the traditional facility. Therefore, a comparison needs a method to quantify the operating expenditure (opex) effect and relate it to current project dollars.

The method will need software that allows quick analysis of alternatives during the initial decision-making phases of the project. Late changes to the project involve rework and therefore may not achieve the expected savings, and late project work typically focuses on capex savings rather than efficiency and uptime.

Additionally, the method requires a framework to ensure commonality of reporting basis to support comparisons.

Full field life cycle

The process design engineer's primary concern is the sizing of equipment and systems in accordance with a composite base case.

The most conservative and simplest approach combines the peak rate of each fluid, though this rarely results in a practical design. More commonly, the engineers will prepare several representative cases from reservoir model production data that include parameters such as high arrival temperature, low arrival temperature, maximum oil flow, maximum gas flow, maximum total liquids, and maximum compressor horsepower at end of life arrival pressure.

While not necessarily a discrete model, the combination of these sizing cases results in a composite overall design base case.

Prediction of what these design cases should be, given time and personnel constraints during this early phase of the project, is difficult. The engineer, therefore, will often select the same definitions as used previously.

Alternatively, a base case for all expected contingencies developed through a year-on-year modeling approach can determine utility use, emissions load, and energy efficiency at the particular combination of production rates provided by the reservoir model.

The software includes a screen for selecting various system capacities to match the composite requirement (Fig. 3).

Fig. 3 shows an example of selecting various system capacities to match this composite requirement. Note that values have a common weight value for comparison purposes only.

In any case, one can evaluate a system requirement on such factors as horsepower, allocated area, direct capex cost, and volumetric throughput from which one can select an appropriate model year or base model updated to change inputs.

For a turbine-driven reinjection gas compressor, one may consider an approach for finding the point at which a 3 × 50% design (two units in operation, one unit spare, 50% of the time) becomes more attractive than a 2 × 100% design.

The 2 × 100% design usually costs less from a capex perspective, due to fewer units, less overall weight, smaller allocated area and simplified piping.

The 3 × 50% design, however, is more efficient in the sense that installed capacity is closer to the requirement for a longer cummulative period. As the facility moves away from the requirement for compression, the 3 × 50% design provides better flexibility with less recycling required.

One needs to quantify the benefits to determine whether this advantage outweighs the capex increase.

In addition to providing additional data during the design phase, another important application of this approach is for providing an ongoing operational target, for energy use or emission, similar to a nameplate processing capacity. Those involved on a day-to-day basis can make significant nondesign-related improvements, such as taking trains offline when not required, driving antisurge or capacity control valves towards minimum position, performing fugitive emission surveys, etc. Real-time data in an indicator or scorecard format could support this effort.6

Establishing current costs

In the early days of the oil industry, associated gas was considered a nuisance and therefore it was vented or flared. As markets developed for using this gas as a fuel or feedstock, the industry started to understand the reservoir drive mechanisms better. It then began using the gas for reinjection as well as for gas sales.

In areas without access to gas markets and below the recoverable threshold for capital-intensive LNG developments, the industry still considers the associated gas as free.

Most producing countries prohibit continuous flaring or will prohibit it in the near future. One valuation, therefore, could be cost avoidance of associated fines.

Assuming that a gas market becomes available due to population center shift, nearby development leading to improved infrastructure, or invention of more efficient small field technologies for transporting the gas to market, an analysis could determine a discounted net present value of the gas based on future sale.

Another method involves looking at the associated cost of using the gas due to emissions taxes. Norway is an example of a country with significant reserves and a mature CO2 tax, currently at about $60/ton CO2. The EU rate for 2008-12 is $33/ton CO2.

Different fuel gases will provide varying amounts of carbon per standard volume, but assuming 133.759 lb CO2/Mscf7 at the EU tax rate, this equates to about $2.21/Mscf. This compares to the 2007 average price for US industrial gas of $7.59/Mscf.8

For valuing the price of carbon emissions separately from the actual cost of the fuel gas itself, the EU rate seems reasonable. Note that Norway's Ministry of Finance is pushing to join the current EU trading system.9

Discounting for present value in countries without any expected requirements in the near future could also be developed.

Alternately, one can use methods provided by several consultancies with a focus towards carbon pricing strategies, such as www.pointcarbon.com.

Model evaluation

A West African FPSO project illustrates the model. This is a common offshore development type and provides some advantages when preparing an illustrative case, as typically storage requirements determine deck area and the deck area is relatively constant regardless of the process configuration.

The base requirement for the facility modeled is 210,000 b/d light sweet crude processing, 360 MMscfd gas processing (used for fuel but primarily sent to reinjection), 90,000 b/d produced-water handling, and 150,000 b/d treated-seawater injection.

Fluid data

The model builds the composite base case and all alternative variants directly from the fluid data. Proper determination of the composite case requirements includes preparing a production depletion curve complete with major power users, such as water injection and gas injection requirements.

The model requires a production depletion curve complete with major power users (Fig. 4).

Notwithstanding potential sensitivity studies based on variation in production rates, Fig. 4 depicts a representative depletion curve tool.

Note that this layout has a quarterly input, which provides sufficient granularity to phase in additional development wells during the early life of the field and for infill wells later on.

Base case preparation

After preparing the model with the fluid data, the engineer must review the peak requirements of the various systems. This involves quantitatively looking at the loads predicted by the software during the field life (Fig. 3).

This screen shows the profile for the generated annual energy production and use (Fig. 5).

The base case model sets the benchmark for comparing other alternatives. This requires that the model generate a set of values for each case.

Once the model updates the base case against the asynchronous peaks determined by the software (Fig. 3), it can produce curves of annual energy production and use against the depletion curve (Fig. 5).

Evaluation metrics of potential interest are the minimum energy intensity (Fig. 6) and CO2/boe (Fig. 7). These provide a measure of the maximum efficiency of the particular design arrangement for a given depletion profile with respect to energy used and CO2 emissions. One can use these results to compare against the capex difference for the various arrangements.

From the emissions perspective, plotting against production provides another way to visualize the drop in efficiency as the facility moves away from the peak throughput point (Fig. 8). The gap between the plots after year 6 provides a measure of the marginal drop in efficiency.

Alternatives studied

The following alternative cases all consider the same depletion profile for illustrative purposes. The peak requirements for the power generation and reinjection gas compression systems are selected based on maximum horsepower. The software model automatically selected the other system peaks on maximum weight basis.

  • Base—FPSO base case.
  • Small—Change 2 × 50% LM2500+ driven reinjection gas compressors to 2 × 50% LM2500 driven reinjection gas compressors (available horsepower reduction, requirement is now 94.5% of site rated available horsepower).
  • Spared—Change 2 × 50% LM2500+ driven reinjection gas compressors to 3 × 50% LM2500+ driven reinjection gas compressors (improve availability).
  • Flexi—Change 2 × 50% LM2500+ driven reinjection gas compressors to 3 × 33% Titan 130 driven reinjection gas compressors (improve availability and late life flexibility).
  • Elec—Change 2 × 50% LM2500+ driven reinjection gas compressors to 2 × 50% variable speed drive (VSD) electric motor driven reinjection gas compressors (power generation forced from 3 × 50% with total load of 18.9 Mw attached to 4 × 33% with total load of 60.8 Mw attached).
  • Import—Change 2 × 50% LM2500+ driven reinjection gas compressors to 2 × 50% VSD electric motor driven reinjection gas compressors and have power generation requirement of 61.5 Mw replaced with import power from shore. Remaining fuel gas users are inert gas generation and low-pressure compression.

Model comparisons

The Elec model shows the best efficiency and emissions performance of the cases studied. A simple net present value calculation for the difference between the Base and Elec cases includes a combination of the fuel used and emissions as modeled, and uses a value of $3.795/MMbtu for fuel (based on 50% of pricing above and 1,000 btu/scf) and $33/ton CO2 for emissions, and annual inflation of 2.5% for all pricing.

This comparison indicates an NPV of $47 million ($31.4 million for fuel gas savings and $15.6 million for emissions avoidance), which can be evaluated against the expected capex increase from the Base model. When combined with expected improvements in uptime, for the modeled case an all-electric drive approach may have significant benefits vs. the traditional design with dedicated gas turbines at each major power user.

Note that the graphs do not include the Spared and Import cases. Spared produces the same efficiency curves as Base. The improvement in availability requires generation of a reliability and maintainability (RAM) study to support the increased capex.

Import provides a much lower local emissions and fuel usage, shifting the burden to shore-based power plant with higher expected efficiency (even with transmission and conversion losses). One needs to balance this against the capex impact of powerline installation, changes in the production facility design and other factors outside the scope of this article.

Further development

Further development work includes:

  • Integrate more sophisticated financial analysis within the model.
  • Combine financials with risks and probabilities of various events such as changing product prices and excessive inflation.
  • Ensure depletion model includes effect of gas cycling in gas lift or gas reinjection situations and water influx for water injection cases.
  • Complete opex model with overhead, uptime, maintenance, other consumables, logistics, modifications and turnarounds, well workovers and decommissioning costs.

Observations

As with any comparison process, one important constraint is to ensure that the model uses a common basis. Additionally, the available data may mask other phenomena. For this reason, below are several potential misunderstandings:

  • Companies can report gas used for fuel on a platform on a boe or heating value basis.10 Depending on the heating value of the fuel gases in use, a boe comparison across fields or used as a benchmark may introduce some confusion. A common conversion of 5.8 Mscf gas to 1 boe depends on the gas calorific value. It uses 1,000 btu/scf. The standard definition of a boe is 5,798,615.481 btu.
  • Btu basis should be standardized because several definitions exist. The article uses the International Steam Unit definition of 1,055.05585262 btu = 1 Joule.
  • Drivers vary by region, so that the engineer should expect alternatives to present different values depending on the development criteria. As an example,11 an onshore Canadian heavy oil development combines high direct energy intensity requirements, a near Arctic environment, and a relatively high market price for gas. These factors would combine to make energy efficiency improvements much more valuable than a similarly sized facility in places such as Venezuela.
  • Companies can report carbon tax in $/ton of carbon or $/ton of CO2. There are 27.3 tons of carbon/100 tons of CO2.
  • If the model can include an agreed basis on a current dollar value to improve percent uptime, one could select a more efficient process despite increased capex and otherwise insufficient opex savings. A RAM study could highlight the uptime differences between the two designs.

References

  1. Vorster, M., "Understand the Difference between Equipment Availability and Utilization," Construction Equipment, Aug. 1, 2007, http://www.constructionequipment.com/article/CA6466790.html.
  2. Upstream Oil and Gas Energy Efficiency, Cambridge Energy Research Associates, unpublished, July 8, 2008.
  3. Energy Use, Loss, and Opportunities Analysis for U.S. Manufacturing and Mining, Energetics Inc. and E3M Inc., December 2004, http://www1.eere.energy.gov/industry/energy_systems/tools.html#publications1.
  4. Jogschies, H., internal Siemens data, unpublished, 2007.
  5. 2008 BP Statistical Review of World Energy, http://www.bp.com/productlanding.do?categoryId=6929&contentId=7044622.
  6. Vanner, R.,"Energy Use in Offshore Oil and Gas Production," September 2005. http://www.psi.org.uk/pdf/Energy Working Paper—June 2005.pdf.
  7. Official Energy Statistics from the US Government, flare gas value, http://www.eia.doe.gov/oiaf/1605/coefficients.html.
  8. Natural Gas Summary, industrial gas 2007 average value, http://tonto.eia.doe.gov/dnav/ng/ng_sum_lsum_dcu_nus_a.htm.
  9. "Norway to join EU carbon-trading scheme," Euractive.com. Mar. 9, 2007. http://www.planetark.org/dailynewsstory.cfm/newsid/40761/story.htm.
  10. Svalheim, S., and King, D., "Life of Field Energy Performance," SPE Paper No. 83993. Offshore Europe, Aberdeen, Sept. 2-5, 2003.
  11. Upstream Oil and Gas energy Efficiency: a Critique of the Siemens Approach, unpublished, Cambridge Energy Research Associates. July 8, 2008.

The author

Theodore J. Mallinson is senior project manager, upstream process/facilities advisor for Siemens Energy Inc. His experience is in process and petroleum engineering, project management, and workflow and business processes. Currently he is program technical lead for redevelopment of the Oil and Gas Manager, a software tool used for design, cost, and weight estimation, and execution plan modeling for upstream oil and gas development projects. Mallinson has a BS in chemical engineering from Rice University, and an MS in petroleum engineering from Heriot-Watt University.