Process design and supply chain optimization of supercritical biodiesel synthesis from waste cooking oils

https://doi.org/10.1016/j.cherd.2013.02.013Get rights and content

Abstract

A small scale biodiesel production facility based on the Mcgyan process is simulated in HYSYS and a follow-up techno-economic analysis is performed. Two feedstocks are analyzed: a soybean oil and waste cooking oil analogs. It is found that the soybean oil based process is not economical at such small scales, whereas the waste oil case has an NPV of $618K with an internal rate of return of 80%. The economic feasibility of a distributed system of small scale biodiesel production facilities in Greater London using waste vegetable cooking feedstock is also investigated. It is found that this system is feasible with a total of 20 installed facilities and an NPV of $1.1MM. A scheme is then implemented which reduces the total capital expenditure per facility based on the mass production of similar facilities. As expected, this scheme reduces the total capital cost of the system. Finally, a Monte Carlo scheme is implemented to study how the variability in economic parameters affects the system. It is found that the system is most sensitive to the sale price of biodiesel but that in all cases a positive NPV is returned. These analyses support the feasibility of small scale locally based biofuel production from locally sourced feestocks.

Highlights

► Small-scale biodiesel production system from waste and soybean oils using Mcgyan process analyzed. ► Distributed production sites in Greater London analyzed. ► System is shown to be profitable over wide range of economic parameters.

Introduction

Biodiesel has been identified as a viable alternative to fossil derived diesel fuel. Biodiesel produced from soybean oil, the industrially established method, can yield more energy than is invested in its production (the biofuel energy content exceeds the fossil energy inputs) by 93% and can reduce (through biological carbon dioxide fixation) the greenhouse gas emissions caused by its production and combustion by up to 41% (Hill et al., 2006). Although environmentally preferable to fossil based fuels, soybean biodiesel costs more than fossil diesel. At 2005 prices, soybean biodiesel production costs are $0.55 per diesel EEL (energy equivalent liter) whereas the wholesale diesel price average was $0.46/L (Hill et al., 2006). Subsidies play an important part in driving the production of biofuels; indeed, in the U.S.A. biodiesel receives a $0.29/EEL federal government subsidy in addition to benefiting from subsidized soybean corps (Camargo et al., 2010).

Soybean oil purchasing can make up to 84% of the total operating costs of a biodiesel production facility (Fortenbery, 2005). It has been suggested that yellow grease could be used instead of soybean oil for the production of biodiesel. Yellow grease is derived from waste vegetable cooking oil and is considerably cheaper than soybean oil – $0.17/lb for yellow grease and $0.33/lb for soybean oil (Fortenbery, 2005).

The conversion of oils to biodiesel is performed using basic homogeneous catalysts to convert oils to fatty acid methyl esters (biodiesel) at moderate conditions (338 K, 1 atm). The base catalyzed process is preferable to the alternative acid catalyzed process as this reaction is slower and higher temperatures are required for similar biodiesel yields ([Freedman and Pryde, 1984], [Vicente et al., 2004], [Vicente et al., 2007]). A saponification side reaction occurs when the traditional biodiesel production methods are applied to feeds high in free fatty acids (such as waste cooking oil). This reduces the biodiesel yield and complicates downstream separation processes. The saponification reaction occurs when free fatty acids are neutralized by the basic catalysts to free fatty acid salts (essentially forming soaps). The Mcgyan process (McNeff et al., 2008) is a patented supercritical process that can avoid the saponification problems by performing the reactions over a proprietary heterogeneous catalyst at supercritical conditions.

Soybeans are a distributed resource and therefore, smaller, distributed biodiesel production facilities may prove to be economically favorable compared to the increasing cost of transportation associated with collecting the soybeans and delivering them to a centralized processing facility. [Fore et al., 2011a], [Fore et al., 2011b] have investigated both the economics and the energy balance of this idea for the production of soybean or canola oil based biodiesel using a yield based conversion efficiency to estimate the production flows. They concluded that the economic viability of these systems is heavily dependent on the price of petroleum diesel and the initial capital expenditure involved.

The principal objective of the work presented here is to investigate whether small scale biodiesel production is feasible and economical. The scale of each process was limited to a feed rate of 20 kg/h of either soybean oil or waste cooking oil. This scale was chosen such that each facility can be trailer mounted to retain the possibility of easy transportation between potential production sites. In order to accomplish this objective, we first simulated (in the HYSYS process simulation software) a Mcgyan process for the production of biodiesel from either virgin soybean oil or waste vegetable cooking oil. Using the results obtained from the HYSYS simulation, we performed a basic techno-economic analysis on each process to calculate the capital and operating costs such that the operating profit could also be calculated. We used the operating profit and capital costs in a discounted cash-flow analysis to find the net present value (NPV) for each case. The HYSYS process simulation results and techno-economic analyses were used as the base case in the investigation of the economic feasibility of a distributed system of small scale biodiesel production facilities based in the Greater London area of the U.K.

Section snippets

Process design

The Mcgyan process is unique among commercial biodiesel processes for its ability to utilize low quality, high free fatty acid content oils as well as the traditional soybean oils. Furthermore, the catalyst used has been shown to catalyze both transesterification and esterification reactions simultaneously (McNeff et al., 2008). In this study, two extreme feedstocks are considered: pure triolein and pure oleic acid (a free fatty acid). Triolein is the triglyceride formed from three oleic acid

Economic analysis

The HYSYS models were used as a base for an economic assessment of the processes. All processes were assumed to run for 8000 h/yr and produce biodiesel that is up to ASTM (American Society for Testing and Materials) specifications. For the economic analyses, the pure triolein case was assumed to be soybean oil and the pure oleic acid case was assumed to be waste oil collected by Sartec Inc. Initially the capital costs of the processes were estimated and the operating costs were found using the

Facility location and supply chain optimization

In this section, we investigate the economic feasibility of a distributed network of biodiesel production facilities installed in the Greater London area of the UK. This location is chosen in order to take advantage of the availability of waste cooking oil. There are over 8000 fast food outlets in Great London (Web-Resource, 2012) as well as numerous other restaurants and office canteens all of which have the potential to be sources of used cooking oil. This location is also used to take

Conclusion

In this paper, we initially developed designs based on the Mcgyan process for the production of biodiesel, using virgin soybean oil and waste cooking oil as the principal feedstock. It was determined that a system operating on a 20 kg/h feed of waste cooking oil has an NPV of $618K based on a 15 year project horizon; whereas, when operating on virgin soybean oil the NPV is −$474K. The analysis shows that the low purchase cost of waste cooking oil ($0.12/lb) is the principal driver of the

Notation

Empty CellDescription
Sets
w{1,,W}demand sources
i  {1, …, I}candidate facilities
j  {1, …, J}set of logicals for capacity learning

Parameters
θdiscount factor (yr)
CAPcapacity (L)
Rrevenue per biodiesel liter ($/yr)
DCdistance based cost ($/mile L)
Dwdemand at source w (L/yr)
OCoperating cost per biodiesel liter ($/L)
Disti,wdistance from i to w (mile)
CAPEXcapital cost of one facility ($)
rdiscount rate (%)
Ttime horizon (yr)

Variables
NPVsystem net present value ($)
Piprofit at facility i ($/yr)
Shipi,wshipment from

Acknowledgments

The authors would like acknowledge the collaboration with Sartec Inc. and in particular Bingwen Yan for his help in the process modeling and economics section. The authors would also like to acknowledge the financial support for this work by a grant from the University of Minnesota Initiative for Renewable Energy and the Environment (large grant RL-0004-09) and by the Abu Dhabi Minnesota Institute for Research Excellence.

References (18)

There are more references available in the full text version of this article.

Cited by (0)

View full text