A. Kister and M. EL-HALWAGI, Texas A&M University, College Station, Texas (U.S.); and C. DUEWALL, Bryan Research & Engineering, Bryan, Texas (U.S.)
Due to the vital role of distillation in hydrocarbon processing and the significant energy consumption associated with distillation systems, there is a critical need to continue to optimize the performance of existing infrastructure and generate insights for new unit designs. Furthermore, it is necessary to enhance the various objectives (e.g., economic, energy, environmental) of these systems. This article proposes a combined approach using exergy loss analysis and product optimization as a rapid method for evaluating and improving existing distillation processes to reduce their costs and environmental impact. Distillation exergy analysis identifies thermodynamic irreversibility and evaluates retrofit possibilities after which the retrofit is economically and environmentally optimized. This two-step process is applied to the deethanizer column of a natural gas liquids (NGLs) fractionator, saving 8,000 tons per year (tpy) of carbon dioxide equivalent (CO2e) emissions and increasing profits by $2 MM.
Distillation stands as the dominant separation technology, bearing responsibility for 90%–95% of all liquid separations spanning across petroleum refineries, chemical processing plants, natural gas purification plants and many more.1 Although highly successful at achieving these desired separations, distillation columns are also known to exhibit very low thermodynamic efficiencies, resulting in the consumption of enormous quantities of energy.2 Estimates suggest that distillation columns account for approximately 40% of the energy used in refineries and 30% of the energy consumed in chemical plants.3 The substantial energy demands of these distillation columns contribute significantly to large CO2 footprints, negatively impacting the sustainability of these processes. Several process troubleshooting, integration and optimization techniques have been proposed to diagnose and improve the performance of distillation systems.4
To address these sub-optimal designs, a two-part method (FIG. 1) was first proposed in Part 1 of this publication to rapidly analyze existing distillation columns for possible improvements. The approach includes: (a) performing an exergy loss analysis to identify regions of excessive thermodynamic irreversibilities within the column, followed by (b) economically and environmentally optimizing the selection of product specifications. Retrofit optimization typically requires discrete variables for decision-making and specialized optimization methods. Exergy analysis simplifies the complex optimization problem into a simple nonlinear programming problem.
Part 1 demonstrated the effectiveness of exergy analysis to identify and rectify thermodynamic irreversibilities in a deethanizer column. Moving the feed stage from 6 to 15 reduced the reboiler and condenser duties significantly, saving $1 MM/yr in utilities and reducing 8,000 metric tpy of CO2e emissions. An analysis of preheating the column feed revealed that while preheat reduces exergy loss, it increases utility cost since refrigeration is considerably more expensive per unit energy than steam heating. This highlights the need for the economic and environmental optimization method detailed here in Part 2 of this series.
Product optimization methodology. After major revamp decisions are made using exergy analysis, product optimization determines the best possible performance. In this case, an objective with economic and environmental factors is chosen to best represent the priorities of the operator. To adequately gauge the column’s performance under different operating conditions, deethanizer product optimization must be done in conjunction with the rest of the columns in the fractionation train. The depropanizer, for example, relies on the deethanizer removing enough ethane that the propane purity specification can be met. For these reasons, the entire train was modeled, with the optimization algorithm manipulating deethanizer performance for each feed to maximize full train performance.
Adapted from literature,13 the overall objective function is train profit minus a penalty for calculated carbon emissions. Profit is calculated using total product revenue RSales plus a $0.02/gal fractionation fee RFee minus feed cost and utilities. CFeed is the total y-grade feed price, CUtilities is the cumulative cost of utilities, and PGWP Penalty is the objective function penalty for CO2e emissions (Eqs. 2–4).
max RSales + RFee – CFeed – CUtilities – PGWP Penalty (2)
CUtilities = Qheat • P gas • + W/Ԑ • P elec (3)
CGWP Penalty = CGWP (Qheat • EFheat + W • EFPow) (4)
Natural gas provides the total required heating utility Qheat, while motors drive pumps and refrigeration compressors using electrical power W. Shown in Eq. 3, the utility cost depends on the prices of natural gas Pgas and electricity Pelec, assumed to be $4.56/MMBtu and $0.08/kWh, respectively. Electric motor efficiency Ԑ is assumed to be 80%.
Carbon penalty, PGWP Penalty in Eq. 4, uses a carbon penalty CGWP of $50/metric ton. This is multiplied by the estimated CO2 the unit directly or indirectly emits. Reboiler heating directly emits 120.2 lb of CO2e per MMBtu of heating (EFheat) according to the U.S. Environmental Protection Agency (EPA), and condenser refrigeration accounts for 832.5 lb of CO2e/MWh of electricity used (EFPow).16 Both emissions factors include CO2, methane (CH4) and nitrous oxide (N2O) emissions. The electrical power emissions factor assumes the electricity was generated by a gas turbine power plant operating at 60% overall efficiency.
Reflux ratio, boil-up ratio and tower pressure were chosen as decision variables to promote solution reliability, consequently affecting the product split. Constraints were set to maintain the ethane purity above 95 vol% and the ratio of ethane-to-propane in the bottoms at 0.05. Exergy analysis was performed with fixed product purities at the constraint value. The fraction flooding was constrained below 100% to prevent flooding.
Product optimization results. With revamp strategies in hand, operators can target relevant economic, environmental and operational criteria for process optimization. The optimization formulation described earlier is applied for each of the three deethanizer feeds (TABLE 5) using a proprietary built-in optimization toola.
The algorithm held the bottoms product C2/C3 ratio at the constrained limit of 0.05. The overhead ethane purity, however, was optimized to above the required specification of 95 vol%. Pushing as much ethane and propane in the bottoms is expected since the bottoms product is more valuable. This change in product split is reflected by the much larger reduction in condenser duty compared to a much smaller reduction in reboiler duty. While increasing ethane purity did increase operating cost and carbon emissions, increased propane production more than offset these additional costs.
In each case, the optimization algorithm raised the tower operating pressure to 353 psig, 333 psig and 331 psig for low-, mid- and high-ethane feeds, respectively. This decreased the relative volatility between the components, leading to a slight increase in total condenser and reboiler duties relative to the exergy analysis results. The benefit of increasing operating pressure is that it raises the bubble point temperature of both products. This slightly increases the sensible heat required to bring the bottoms product up to the bubble point, while also slightly reducing the refrigeration required to cool the overhead product to its dewpoint. This allows the refrigeration utility to be produced at a higher temperature, greatly reducing the cost of refrigeration. In each case, heating costs remain fixed at $4.56/MMBtu while the refrigeration cost was calculated by simulation, averaging $12.8/MMBtu. Not only did raising the tower pressure shift the load towards the lower cost utility, but it also lowered the cost of refrigeration itself by $0.20/MMBtu–$0.80/MMBtu. This relationship between tower pressure and refrigeration cost for each of the three feed streams is shown in FIG. 8.
The effects of the exergy analysis and subsequent optimization on reboiler and condenser duties, deethanizer GWP-100 and overall train profit can be seen in FIGS. 9 and 10. The exergy analysis is shown to reduce reboiler and condenser duties as well as carbon emissions while raising profits to a limited extent for all three feed cases. The optimization further increases profit margins while conceding minor increases in utility consumption and carbon emissions.
For the mid- and high-ethane feeds, the optimization algorithm increased the GWP-100 emissions relative to the exergy analysis case by 1,400 metric tpy and 1,600 metric tpy of CO2e emissions. While the optimized emissions were still lower than the base case, these results suggest that the increased product value outweighs the effects of the carbon penalty. Although product optimization incurred a slight penalty to the environmental component of the objective function, it doubled the profitability improvement of the exergy analysis component. Annual profit increased by an additional $1 MM, totaling $2 MM or a 16% increase from the base case.
Takeaways. In this two-part publication, a combined approach for analyzing and optimizing hydrocarbon distillation columns to enhance sustainability and cost-effectiveness has been proposed. The first part involves performing an exergy loss analysis using a commercial simulator. This can identify thermodynamic irreversibilities within the column which may be reduced through the application of targeted revamp strategies. After revamp, the new facility is economically and environmentally optimized. When combined, this methodology can suggest comprehensive and insightful improvements to a distillation column, with the ability to lower CO2 emissions and save on cost.
This method was applied to a case study involving a deethanizer in an NGL fractionation train. The exergy analysis revealed that lowering the feed stage from 6 to 15 would reduce the reboiler and condenser duties by 9% and 19%, respectively, reducing emissions by up to 8,000 metric tpy of CO2e (8%) and increasing profits after carbon penalty by up to $1 MM/yr (8%). Following this, the optimization was performed, successfully identifying the most optimal operating points. These efforts uncovered an additional $1 MM/yr, improving profit by a total of $2 MM/yr (16%) from the original operating point.
The results presented in this article show that this one-two punch method can be highly successful in quickly and methodically identifying major opportunities within existing distillation columns for environmental and economic improvements. The proposed approach can be evolved to address a wide variety of applications in the hydrocarbon processing sector. HP
NOTE
Bryan Research & Engineering’s ProMax® Optimization Tool
LITERATURE CITED
Agrawal, R. and G. R. Tumbalam G. R., “Misconceptions about efficiency and maturity of distillation,” AIChE Journal, 2020, online: https://doi.org/10.1002/aic.16294
Sholl, D. S. and R. P. Lively, , “Seven chemical separations to change the world,” Nature, April 2016, online: https://www.nature.com/articles/532435a
Oak Ridge National Laboratory & BCS Inc., “Materials for separation technologies: Energy and emission reduction opportunities,” 2005, online: https://doi.org/10.2172/1218755
Kister, H. Z., “Distillation diagnostics: An engineer's guidebook,” 2025, John Wiley & Sons.
Kiss, A. A., “Distillation technology: Still young and full of breakthrough opportunities,” Journal of Chemical Technology & Biotechnology, 2014, online: https://doi.org/10.1002/jctb.4262
Al-Douri, A., D. Sengupta and M. M. El-Halwagi, “Shale gas monetization: A review of downstream processing to chemicals and fuels,” Journal of Natural Gas Science & Engineering, 2017, online: https://www.sciencedirect.com/science/article/abs/pii/S1875510017302299
International Energy Agency (IEA), “World energy outlook 2012,” 2012, online: https://www.iea.org/reports/world-energy-outlook-2012
Shingan, B., N. Banerjee, M. Kumawat, P. Mitra, V. Parthasarthy V. and Singh, “Liquefied natural gas value chain: A comprehensive review and analysis,” Chemical Engineering & Technology, 2024, online: https://doi.org/10.1002/ceat.202300304
Islam, M., S. A. Al-Sobhi, A. Naquash, M. A. Qyyum and M. Lee, “Optimal process selection for natural gas liquids recovery: Energy, exergy, economic, and environmental perspectives,” Energy, 2024, online: https://doi.org/10.1016/j.energy.2023.129757
Asani, R. R., R. Mukherjee and M. M. El-Halwagi,, “Optimal selection of shale gas processing and NGL recovery plant from multiperiod simulation,” Process Integration and Optimization for Sustainability, 2021, online: https://doi.org/10.1007/s41660-020-00140-9
Soares Pinto, F., R. Zemp, M. Jobson and R. Smith, “Thermodynamic optimization of distillation columns,” Chemical Engineering Science, 2011, online: https://doi.org/10.1016/j.ces.2011.03.022
Zemp, R., “Thermodynamic analysis of separation systems (Thesis),” University of Manchester, 1994.
Duewall, C. and M. M. El-Halwagi, “Multi-objective Bayesian optimization of NGL fractionators using process simulation,” online: https://ssrn.com/abstract=4912537
El-Halwagi, M. M., “Sustainable design through process integration: Fundamentals and applications to industrial pollution prevention, resource conservation, and profitability enhancement,”
Felbab, N., B. Patel, M. El-Halwagi, D. Hildebrandt and D. Glasser D., “Vapor recompression for efficient distillation: A new synthesis perspective on standard configurations,” AIChE Journal, 2013, online: https://aiche.onlinelibrary.wiley.com/doi/10.1002/aic.14070
U.S. Environmental Protection Agency (EPA) AP-42, Vol. 1, Chapter 1.4, “Natural gas-fired heaters,” 1996.
Avi Kister is a Senior Associate Process Engineer at Celanese, where he supports the company's methanol and carbon capture operations. Avi earned a BS degree in chemical engineering from the University of California, Los Angeles (U.S.) in 2022 and an MS degree in chemical engineering from Texas A&M University in 2024. His graduate thesis was titled "Improving sustainability of existing distillation columns through exergy analysis and revamp techniques." Kister has previously published and presented work on parting box technology and plans to continue his involvement in advancing distillation technology.
Mahmoud El-Halwagi holds the Bryan Research & Engineering Chair in the Artie McFerrin Department of Chemical Engineering at Texas A&M University. Dr. El-Halwagi is the leader of the Process Integration and Systems Optimization group, focusing on sustainable process synthesis, design, operation, integration and optimization. The theme of his group’s research is the development of systematic methodologies that enable chemical engineers to identify optimum, sustainable and creative strategies that lead to productivity enhancement, yield improvement, debottlenecking, pollution prevention and energy conservation.
Clinton Duewall is the Applications Engineering Manager at Bryan Research & Engineering. His team is responsible for creating, automating and integrating ProMax simulations as first-principles digital twins and online optimizers. Dr. Duewall earned a BS degree and PhD in chemical engineering from Texas A&M University, and is an Engineer-In-Training in the state of Texas.