K. WELLS and K. JENNETT, Veolia Water Technologies & Solutions, Philadelphia, Pennsylvania (U.S.)
Improving energy efficiency is an ongoing objective for refineries to benefit operability, throughput and profitability. It is acknowledged in the field that somewhere between 5% and 38% of the energy input to refining petroleum is wasted.1 Recently, additional challenges of rising energy costs and greenhouse gas (GHG) emissions reduction incentives have increased the drive to capture energy efficiency opportunities.
These twin challenges can be partially satisfied with energy conservation projects and process modifications. To be sure, many of these efforts will be capital intensive with long lead times. However, immediate steps are available that can show results within weeks. Justifying these actions requires an understanding of energy sources, energy costs, fuel carbon content and the interactions of fuel and steam balances unique to each refinery.
The magnitude of driving forces provided by these challenges varies in different regions of the world, but the overall trend is increasing. Using the U.S. as an example, natural gas costs have more than tripled over the past year and are expected to stay above $7/MMBtu. At the same time, most energy companies have stated goals of reducing CO2 emissions.
This article will examine the drivers of change, promote the understanding of the energy consumption of an “average†refinery, and explore five low-capital projects to improve refinery profitability and reduce carbon footprint.
DRIVING FORCES OF CHANGE
Price. North America enjoyed very low natural gas prices for more than a decade, primarily due to hydraulic fracturing technology brought online in the early 2000s that enabled economically viable access to more natural gas reserves than previously possible. More recently, low demand in the early days of the pandemic pushed these prices even lower.
However, two factors have pushed natural gas prices dramatically upward. First, post-pandemic economic expansion has increased demand. Typically, drilling activity that was nearly halted in 2020 should respond within 6 mos to the higher prices, increasing supply and normalizing price. However, recent geopolitical crises mean that exports of liquefied natural gas (LNG) to Europe will be maximized. This supports a higher price, probably around $7/MMBtu, for the foreseeable future (FIG. 1).
Note: On a per Btu basis, natural gas is always far less expensive than crude or liquid fuels. For instance, on April 26, 2022, natural gas was priced at $6.89/MMBtu wholesale; by contrast, WTI Crude was $17.17/MMBtu, or a 250% premium per Btu ($99.60/bbl and 5.8 MMBtu/bbl).2 This price discrepancy is due to engineering and logistical factors, which had previously limited natural gas to fixed locations on a pipeline network—while this increased supply, the fixed customer base kept prices low. This is changing rapidly, with LNG exports growing 50% in 2021, even before the current crisis.
Natural gas is also used as a feedstock to make hydrogen. While that topic is not covered here, economic analyses should be careful to exclude that consumption from the energy bill.
GHG reduction. This parameter is featured prominently in the reports of every U.S. refining company. GHG measurement includes three scopes:
It is estimated that Scope 1 accounts for only 5%–10% of total emissions for fuel refineries. Scopes 1, 2 and 3 combined have sometimes been referred to as “well to wheels†in the petroleum space.
Most U.S. petroleum refineries have promised to drastically reduce GHG intensity in the future. These pledges represent at least 75% of the barrels processed and include all top five refiners. Based on each refiner’s website, their pledges fall into two timeframes: 50% reduction by 2030 and/or 100% reduction by 2050. Presently, no financial penalty has been established for missing these pledges.
U.S. federal actions on industrially generated GHG are uncertain. In the meantime, state-by-state goals are being phased in for states that contain about 33% of U.S. refining capacity. So far, only California has an actual cost on marginal CO2, presently around $30/MMt.1 This potentially adds about 50% to the energy costs for the average refinery. The goals in the rest of the states are aspirational, with no current financial penalty. The U.S. state-by-state GHG reduction targets shown in TABLE 1 were gathered from The Center for Climate and Energy Solutions,4 and refinery capacity data is from the EIA.5
“Typical†refinery energy facts. Of course, every refinery has a unique energy usage and carbon footprint: there is no such thing as a “typical†refinery. Nationally published data from The Office of Energy Efficiency & Renewable Energy (EERE),3 has been proportioned to a “typical†100,000-bpd refinery. These numbers will be used here as examples, but specific refinery data should be used to evaluate projects.
Temperature hill—Every refinery process occurs at elevated temperatures. The duty (heat) to achieve these temperatures comes from process furnaces. Fired furnace duty is minimized by recovering the heat of hot products to preheat the feed, typically in shell-and-tube heat exchangers. The design and cleanliness of these heat exchanger networks are important to how much heat is captured and how much is wasted. It is estimated that the typical refinery recovers 33%–66% of its waste heat in this manner. The rest is rejected to air or cooling water.
Steam—Refineries also require a significant amount of steam, around 110 lb/bbl processed. This varies markedly depending on refinery configuration, especially the internal electricity generation. The refiner may use this steam directly in the process, or to drive pump and compressor turbines. Some of this steam is generated by waste heat from the processes, and some by fired boilers. Notoriously, steam imbalances often result in refineries throwing away (venting) low-pressure (LP) steam. This may reduce the benefit of saving energy in these systems.
Efficiency—Of course, not all energy expended is actually absorbed by the process fluid or otherwise converted to useful work. Refiners, in the case of a furnace or boiler, consider the ratio of absorbed duty to fired duty. This is called the “furnace efficiency,†and colloquially that varies from 80%–95%. By contrast, according to the EERE, refinery efficiency (the percentage of primary energy/applied energy) is 62%. Obviously, this discrepancy is because refinery engineers have a much narrower definition of “efficiency,†one that neglects low-value waste heat rejected to cooling water or atmosphere—it is estimated that between 5% and 38% of the energy input to refining petroleum is wasted. Clearly, economic and engineering reasons exist for some of this waste. However, the projects listed in the following section show there is also fertile ground for energy and GHG savings, given the current situation.
Gross energy consumption—In our “typical†100,000-bpd refinery, the energy consumption is 2,200 MMBtu/hr. For the fuel mix below, the refinery energy bill would be ~$65 MM/yr, approximately doubled since early 2021. About two-thirds of refinery energy consumption involves distillation or steam generation, rather than driving chemical reactions.
Refinery carbon footprint—That same 100,000-bpd refinery would generate ~73 lb of CO2/bbl processed, or about 1.2 Btpy. This depends on both the amount of energy and the percentage that comes from fluid catalytic cracking (FCC) coke.
Refinery energy sources—Refineries get their energy from a variety of sources.5 Based on the EERE, those sources include:
TABLE 2 and FIG. 2 summarize the energy picture of the “average†refinery.
EXAMPLE PROJECT IMPACTS
Five low-capital expenditure projects have been listed in the following section, prioritized by certainty of the data typically available. These savings are all based on minimizing wasted heat, either by reducing hot out streams or by increasing heat recovery from those products. The value of water and chemical consumption, product yield or throughput improvements have not been included in these calculations. Often, these could be many times higher, but they are beyond the scope here. Individual fuel carbon contents are shown in literature.7
Steam condensate recovery. Condensate recovery is a process to reuse the water and sensible heat contained in the discharged condensate. Recovering condensate can lead to significant savings of energy, make-up water and chemical treatment, as well as improve working conditions and reduce a plant's carbon footprint. For a 500,000-lb/hr steam production, returning an additional 10% of hot condensate to use as boiler feedwater will provide energy savings of 15 MMBtu/hr, which equals a reduction of $1.1 MM/yr in external fuel costs and 8,200 metric t of CO2. There are many ways to increase savings related to condensate: identifying steam leaks and condensate losses that are easily captured and can be re-used as boiler feedwater; a robust condensate monitoring program and advanced treatment solutions can minimize these leaks and reduce maintenance costs; and working with the water treatment supplier to quantify potential energy, water and chemical savings.
Boiler cycles. Steam boilers present a range of opportunities for process heating enhancements; however, perhaps the easiest way to impact fuel consumption and expenditures is through maximizing the cycles of concentration (reducing the blowdown). This change can bring significant improvements in terms of economies of water, chemistry and fuel. Reducing the blowdown by 2% can represent major energy savings annually. A variety of options are available to increase boiler cycles that include water source control, external and internal chemical treatment changes, re-evaluation of steam purity requirements, and automation for more precise control. For that same 500,000-lb/hr steam system, reducing the blowdown from 8% (12 cycles) to 6% (16 cycles) would have an energy savings of 4 MMBtu/hr, equaling $290,000/yr. Again, consider working with the chemical supplier or water treater to quantify the savings in reducing boiler blowdown.
FCC slurry fouling. The slurry loop is a recirculating liquid heavy hydrocarbon stream from the bottom of the FCCU main fractionator tower that condenses the 1,000°F gaseous products of the cracking reaction, cooling them to 700°F by exchanging heat against FCC fresh feed and waste heat steam generators. Many refineries also use slurry loop heat to drive product gasoline debutanizer reboilers. As the fresh feed exchangers foul, more duty is transferred to steam instead of recovering this heat into FCC feed preheat on the cold side of the fresh feed exchangers. This duty is made up in the feed furnace, or by burning more coke in the reaction loop. While coke is “free†from an energy cost perspective, it adds to the CO2 load of the refinery. FCC slurry bundles can be cleaned online, perhaps without process penalty during colder weather. However, slurry loops foul quickly, with a typical cleaning cycle lasting just 3 mos. Therefore, chemical treatment to reduce cleaning frequency or protect the performance of critical reboilers is often justified. In the hypothetical average refinery, the slurry loop transfers 20 MMBtu/hr to the fresh feed when clean, but just 4 MMBtu/hr when fouled. As stated earlier, much of this heat is recovered in the steam generators, but about half of the steam produced from this process ends up being vented to the atmosphere due to steam balance issues, ultimately resulting in a loss of both energy and water. Further, the slurry heat not transferred to the feed is instead made up by coke burn. A well-managed antifoulant and cleaning program, in partnership with the chemical supplier, might save an average of 6 MMBtu/hr (0.3%) of refinery heat duty, after accounting for partial energy recovery in the steam generators. This is $433,000/yr, or 0.7% of the energy bill. Importantly, because the replacement fuel (FCC coke) is almost all carbon, this one change can reduce the refinery carbon footprint by 0.4%. Keeping to the topic of energy costs and GHGs, the effects on yield and throughput from the slurry loop fouling—which are typically many times greater—have been neglected here.
Ultra-low sulfur diesel (ULSD) fouling. Hydrotreater reactions take place at 650°F, and a 24,000-bpd ULSD unit has cold feed exchangers (CFEs) that recover 100 MMBtu/hr when freshly cleaned. Unfortunately, these exchangers get dirty in normal use. This duty might fall by 30% when fouled. The obvious choice is to clean these exchangers to recover the duty to the exchangers, reducing the load on the furnace. Calculating the value of mechanically cleaning these exchangers presents two challenges. First, the exchangers will start fouling again as soon as they are put back in service, so the energy savings only apply to a given period. In addition, many hydrotreaters do not have a provision for bypassing while the unit is running, even at reduced rates. This means cleaning must be conducted during turnarounds, or at least when an opportunity to shut down arises. These two challenges point to chemical antifoulants, which will slow the decline to fully fouled conditions and reduce the frequency of mechanical cleanings. While a detailed duty cleaning cycle calculation is necessary, for the purposes of this discussion, a 67% reduction in fouling (or a savings of 20 MMBtu/hr) is assumed. This represents a savings of $1.4 MM/yr on energy cost (more than 2% of refinery totals), and a reduction of 11,000 metric t of CO2, almost 1% of the whole refinery.
Crude preheat fouling. The biggest single consumer of refinery energy is the crude unit: it must heat up the entire refinery feed from ambient temperature to about 650°F while vaporizing a large portion of it. The crude preheat network in the hypothetical refinery recovers 150 MMBtu/hr from hot tower products when clean but just 75 MMBtu/hr when fully fouled. This lost duty can typically, but not always, be made up in the crude furnace, albeit at a high expense in cost and carbon emissions. The ability to do this is constrained by such things as environmental limits and tube skin temperatures. Crude unit preheat bundles have a longer cleaning cycle of roughly ≥ 1 yr. They can usually be cleaned online, subject to the heat removal needs of the crude tower temperature controls. An additional complication of calculating the impact of cleaning is the “network effect.†Cold crude on its way to the furnace exchanges heat from each tower product in succession, starting with the coolest (naphtha: 250°F) and ending with the hottest (vacuum bottoms: 700°F). Heat gained by cleaning exchangers early in the network reduces the driving forces downstream, attenuating the duty gain from early exchangers by 50% or more. Sometimes, pressure drop rather than heat transfer can be an important constraint. Another issue is that the crude unit sees much more feed variation than any other unit in the refinery. All these factors make modeling the crude preheat network among the most complicated predictive tasks in a refinery. For the purposes of this discussion, it is assumed that a robust, optimized exchanger cleaning schedule, along with chemical treatment, can, on average, reduce the fouling losses by 67%, from 75 MMBtu/hr down to 25 MMBtu/hr. This represents $3.6 MM of the external energy bill and 2.3% of the carbon footprint. Unsurprisingly, the crude unit is the biggest source of energy and GHG savings.
TABLE 3 is a summary of the low-cost capital projects.
Takeaways. The driving forces for change, in both energy costs and GHG emissions, have been examined, and five low-capital expenditure examples to move towards those goals have been highlighted. These projects reduce external energy costs by 10% and GHG emissions by 5%. Because the crude unit consumes the most energy in the refinery, the improvement potential is greatest there. However, significant savings are possible in downstream units and utilities, with less uncertainty in the payout calculation. Finally, it should be emphasized that many projects that would provide similar or greater savings are outside the authors’ expertise.
According to the U.S. Department of Energy, refineries waste 38.7% of the energy input needed to turn crude oil into useful, environmentally friendly products (1,444 TBtu lost from 3,728 TBtu input).5 Colloquially, refinery process engineers view efficiency more narrowly, as 5%–20% lost. Regardless, in view of rising natural gas prices and the importance of reducing carbon footprint, any loss is a sufficient argument to work towards sustainability improvements while also providing strong return on investment. HP
LITERATURE CITED
KENNETH WELLS is a Product Applications Engineer for Veolia Water Technologies & Solutions with 41 yr of refinery experience. Wells specializes in antifoulant applications but has also worked on desalting, corrosion control, boiler and cooling tower refinery applications throughout North America. He also specializes in the economic justification of chemical treatment. Wells earned a BS degree in chemical engineering from the University of Delaware.
KENDALL JENNETT is an Account Manager for Veolia Water Technologies & Solutions and her experience stems from more than 10 yr in the refining industry, specializing in desalting and hydroprocessing applications. Jennet earned a BS degree in chemical engineering from Drexel University in Pennsylvania.
