(518b) Optimization of Energy Supply Chain in Upstream Petroleum Operations Under Uncertainty

In 2009 the total upstream petroleum operation cost (finding plus lifting) in Central and South America was about 26.64 US $ per BOE [1]. In the case of ECOPETROL S.A, energy consumption is about 23 % of the total upstream operation cost. In addition to that, energy consumption from oil fields is almost 80% of the total energy demand and oil treatment facilities consume the last 20% in the oil production supply chain. Usually, the resources used to generate energy are available on site and are typically the cheapest source of energy. Moreover, from the supply side, self-generation from natural gas plants provide 90% of the energy required and 10 % of the energy consumption is supplied from the grid [2]. However, for planning long range production facilities, the energy supply choices (self-generation, cogeneration, supply from the grid, gas pipelines and electricity transmission lines, etc.) are subject to uncertainty in electricity (from the grid) prices, production plans (and thus energy demand) and primary energy availability on site (i.e. gas and crude, renewables and biomass). Thus, energy consumption in upstream petroleum operations is an important issue, and optimization tools are needed in order to support decision-making processes.   

            In this work, an optimization model for the planning of the energy supply chain in upstream operations under uncertainty is presented.  The main objective of the supply chain model was to optimize the expected net present value (E[NPV]) taking into account uncertainty in electricity prices, natural gas availability on site and energy demand. The optimization problem was first formulated as a two-stage stochastic programming MILP model (with 600 scenarios) and then solved using the sample average approximation method on an Excel-GAMS interface, the solver CPLEX (12.1) was used to solve the optimization problems. 154 optimal designs (with different economic performance) for the energy supply chain were identified, and risk curves were constructed for each of them. Different risk curves of interest for decision makers were identified: one with maximum expected NPV, deterministic solution, risk-averse and risk taker design. Finally, probabilistic metrics like value at risk (VaR), opportunity value (OV) and risk area ratio (RAR) for financial risk management were applied to the risk curves of interest.

[1]  http://www.eia.gov/tools/faqs/faq.cfm?id=367&t=6

[2] IPIECA. 2007. “The global oil and gas industry association for environmental and social issues.” (http://www.ipieca.org/sites/default/files/publications/Saving_Energy.pdf).