2009 Life Cycle Analysis

Energy Balance for Piedmont Biofuels in the Production of Biodiesel

Using Rendered Chicken Fat for 2008

Collaborative research of: North Carolina State University, Masters of Microbial

Biotechnology Program; R. Burton et al. Piedmont Biofuels Industrial. And NC State

University: Dr.S.D. Terry (Mechanical and Aerospace Engineering Department), Dr. A.

Hobbs (NCSU Solar House), & Dr. M. Flickinger (BTEC).

23 April 2009

Abstract: Piedmont Biofuels, a North Carolina based biodiesel production plant,

produced 1,046,389 gallons of biodiesel in 2008 through the transesterification process

of chicken fat. Piedmont Biofuels was interested in looking at their energy balance for

2008 in order to improve the efficiency of their biodiesel production process. This study

found that in 2008, Piedmont Biofuels had a net energy ratio (NER) of 2.54. The energy

balance model used in this study was compared to other energy balance models such as

the NREL and GREET. In this paper, we have presented the results of these energy

balance analyses and offered suggestions to Piedmont Biofuels on possible ways to

improve their overall net energy ratio.

1-Introduction

Increasing concerns about global warming and energy consumption has

stimulated efforts for the production of cleaner, eco-friendly, renewable energy fuels.

Piedmont Biofuels, a worker and member owned cooperative located in Pittsboro, North

Carolina, promotes, produces, and sells biodiesel fuel and was established with this goal

in mind. In 2008, Piedmont Biofuels used 1,268,241 gallons of rendered chicken fat to

produce 1,046,389 gallons of biodiesel. To produce biodiesel, Piedmont Biofuels uses a

transesterification process, in which methanol reacts with the triglyceride oils contained

in rendered chicken fats to form fatty acid methyl esters (biodiesel) and glycerin. In

addition, heat and a strong base catalyst such as sodium hydroxide or potassium

hydroxide are required for this reaction. However, Piedmont Biofuels uses potassium

methylate (methanol + potassium hydroxide) as the catalyst for this reaction. Since

economic viability is a major concern in biodiesel production, it is important to adopt a

biorefinery model to improve the production process. This mainly calls for the recovery

and purification of the process waste to minimize the loss of recyclable materials. In early

2009, Piedmont Biofuels began implementing changes in their production process to

include a biorefinery model in their plant [See Figures: 1 & 2], which includes methanol

and biodiesel recovery from process waste and glycerin refining. In addition, Piedmont

Biofuels added wash water acidulation and biodiesel refining to their process.

The energy balance for a biofuel production system can be defined as the relation

between the energy produced (output/ kg biodiesel) and the energy consumed (input/kg

biodiesel) for each unit product, thus making it an important index for the economic and

environmental feasibility of a biofuel project (da Costa, 2006). In order to improve and

access the biodiesel production with biorefinery model, Piedmont Biofuels wanted to

know their energy balance for 2008, thus allowing them to compare it with 2009, when

they began implementing changes to their process.

The following four groups of graduate students from North Carolina State

Universitys Masters in Microbial Biotechnology worked with Piedmont Biofuels: the

Energy Balance Group worked on an energy balance study on the Piedmont Biofuels

Industrial plant for all of its processes relating to biodiesel production, the Sidestreams

Group conducted feasibility studies for alternative techniques for dealing with

sidestreams including wash water and glycerin, the Enzymatic Lab Group tested the

effectiveness of Novozymes (a Denmark based biotechnology company with a plant

located in Franklinton, NC which is a major producer of enzymes) enzymes for the

production of biodiesel from triglycerides and free fatty acids under various conditions,

while the SBIR Grant Group  wrote  a SBIR grant to fund enzymatic biodiesel production

research/pilot plant. Our group, the Energy Balance Group, was given the task of

performing an energy balance study for Piedmont Biofuels production process.

For the basis of our energy balance model, we used a customized version of the

NREL model, which was expanded to include human labor and the biorefinery process.

We compared our model to the NREL model, the GREET model, and a modified GREET

model which included human labor. According to the model of Sheehan et al. (1998),

also known as the National Renewable Energy Laboratory (NREL) study, the energy

ratio should be defined as: RNREL = Eb/E.f2, where Eb is the total input energy for the

biodiesel and E.f2 is the fraction of the input energy associated with biodiesel. The

GREET model calculates fossil fuel energy requirements on a per-mile basis for the

major alternative fuels and vehicle systems (Andress, 2002). Both the NREL and GREET

models assigned the energy in proportion to the mass fraction of output (Pradhan, 2007).

The purpose of this study was to write an energy balance study on the Piedmont

Biofuels Industrial plant for all of its processes relating to biodiesel production. This

would allow Piedmont Biofuels to make a comparison of the energy balance for 2008 to

the energy balance for 2009. Piedmont Biofuels introduced different methods in 2009 to

their production process in an attempt to improve the energy balance.

2-Methodology

The study included all upstream energy consumption, energy used in transporting

and processing the process ingredients, and the energy used in biodiesel production. To

calculate the energy balance we included both primary and secondary process energy.

Primary energy is the energy content of the feedstock, methanol, and methoxide (Table

1). Secondary energy input for the biodiesel production was the total energy used in the

transportation of methanol, potassium methylate, wash water, and feedstock. This mainly

included the fuel type used and distance traveled during transportation. Electricity

consumption was also included in secondary energy (Table 3 & Figure 3). Electricity

consumption of the entire plant was monitored to check the efficiency of the pumps and

motors used in the processes from raw materials to biodiesel production. Also, Piedmont

Biofuels used solar panels to provide energy for methanol recovery and feedstock heating

(Table 4). Solar energy is a renewable source of energy, thus it was not included in the

energy balance calculation as an energy input. In addition, labor was included as

secondary energy input. A typical energy input per worker is 100 W; we calculated labor

input based on 360kJ/ hour (Energy input labeling standards, 2008). All the calculations

were done by converting the energy consumptions into BTUs. Mass energy balance was

calculated to determine the mass losses during the production stages (Table 4).

A Net Energy Ratio (NER) was calculated for the process to provide a measure

for how much fossil fuel equivalents were used to produce the fuel, compared to the total

energy value of the biodiesel. This yields a measure for the renewability of the process.

The definition for the NER, NER=Eb/ (E1*f1+E2*f2+E3*f3), was adopted from the 1998

NREL study (Sheehan et al. 1998). For this definition, Eb is the energy value of the

biodiesel. E1 equals the energy input required to grow the raw feedstock, with f1 being the

fraction of that energy allocated to biodiesel. E2 is the energy used in processing the raw

materials before they are fed to chickens as chicken feed, with f2 being the fraction of that

energy allocated to the feedstock. E3 equals the energy used in the transesterification

process; and f3 again being the fraction allocated to biodiesel.

A feed conversion ratio (FCR) was used to convert the energy allocated to initial

raw materials used as chicken feed to the energy in the fat. The FCR was assumed to be

1.88 lb feed per 1 lb chicken meat for a young chicken (<40 days) (J. Petitte, personal

correspondence). In other words, if total energy required to produce 1 lb of chicken feed

was x BTUs, 1 lb of chicken fat created from that feed would require 1.88x BTUs of

energy.

3- Data

The presented data corresponds to the data provided by Piedmont Biofuels, and

information that was gathered during the evaluation of Piedmont Biofuels process.  The

study found the energy balance for Piedmont Biofuels to be 2.54, using the previously

described, customized version of the NREL model. Other models such as the GREET +

labor, GREET, and NREL models calculated a NER of 2.87, 3.00 and 3.38, respectively

(Table 4).

Figure 1: 2008 Biodiesel Production Process

In Figure 1, above, the 2008 biodiesel production process is given in a simplified

form. Methanol and methoxide are mixed together, and are then combined with chicken

fat in the reaction vessel. After approximately 4 hours the reaction is complete and is

allowed to settle for separation of the biodiesel and glycerin phases. Next, the biodiesel is

washed to remove methanol and other contaminants, and the glycerin is disposed of

together with the wash water. Solar heating provided about 5% of the power used in the

plant.

This process differs from the 2009 process in the following ways: in 2009 a

biorefinery module was added that is able to remove most of the excess methanol from

the biodiesel and glycerin phases and can therefore recycle this energy-rich component.

After removal of the methanol present in the glycerin phase it is more concentrated (65%

glycerol instead of 35%) and can potentially be used for other purposes. This process is

presented in graphical form in Figure 2.

Figure 2: 2009 Biodiesel Production Process

In Figure 3, below, the electricity use is presented in graphical form. As can be

seen from the figure, the largest energy users are the various resistance heaters present in

the process to heat the large reaction and wash vessels. Potentially this could be an area

where Piedmont Biofuels could save energy. Furthermore, the large pumps, e.g., the

vacuum pump and the air compressor in biorefinery process use a fair amount of

electricity.

Figure 3: Electricity Consumption: This chart lists the main electricity utilizations in the process.

Farming inputs Btu/Gal

Chicken fat 17968.8 11622.7 15813.8

Soybean oil 12529.6 4666.9 9621.0

Model Our Model NREL GREET

Table 1: Energy input in Feedstocks. This table compares the energy inputs from farming and

processing the feedstock between the author’s model and other commonly cited studies such as

NREL and GREET.

Table 1, above, lists the different farm and feedstock processing energy inputs for

both soybeans and chicken fat. Depending on the model and assumptions used, the

energy inputs are vastly different. This is due to significant differences in the various

energy inputs for each step in the study, as well as, different inclusion criteria. These

differences are analyzed in greater detail in Pradhan, et al., 2007.  The fertilizer energy

input used in this model is broken down in Table 2 for each of the elements that make up

the mixture.

Fertilizer Fertilizer Energy for Soybeans

Nitrogen 2822.64 BTU/bushel

Phosphate 2473.80 BTU/bushel

Potassium 2736.50 BTU/bushel

Total 8032.94 BTU/bushel

Table 2: Energy for Fertilizer processes to Produce Soybeans. This table shows the approximate

energy required to produce the fertilizer used in growing 1 bushel of soybeans.

A more detailed analysis of the electricity usages in the biodiesel production plant

is given in Table 3. This is a condensed table that only shows the major energy draws in

the system, as there are dozens of total appliances in the system. As can be seen in the

table, the wash tank heaters, feedstock tank heater and reactor heaters use a majority of

the total electricity. Other large draws in the system are the air compressors and the

vacuum pump used in the biorefinery process.

Table 3: Electricity Consumption. Table 3 shows the largest electrical energy usages in the process.

The main energy expenditures in this area can be found in the heating elements and the larger

pumps.

Table 4: Energy Balance of Various Models. This table shows the composition of the NER for

different starting assumptions, and provides the total value of the NER for each scenario. Solar

heating values are not included in the NER and are only shown to illustrate how much electrical

energy is saved.

Table 4 shows the result of the energy balance for this study. It uses the total

energy inputs for 2008, together with the total amount of biodiesel produced in that year,

to derive a relative NER based on a gallon of biodiesel. Besides the model used in this

study, several other models are given to illustrate some of the assumptions made about

the input energies. The GREET model is fairly close to the modified model in terms of

farming and processing inputs. The main difference is that it does not include the waste

disposal energy costs associated with disposing of glycerin and wash water.

BTU/Gallon Biodiesel

GREET GREET w/ Labor

E1f1 15814 15814

E2f2 6842 6842

E3f3 24382 24399

Eb 125100 125100

NER 2.66 2.66

Table 5:  Energy in farming processes. This table shows the energy required to produce biodiesel

from chicken fat if the assumed energy inputs from the GREET model were converted to chicken fat

biodiesel.

Table 5: Legend for Energy in farming processes

The purpose of Table 5 is to illustrate the difference inclusion of labor energy

inputs makes on the NER. As can be seen from the final values, it is a minor amount in

comparison to several other inputs and does not change the ratio at all. The energy used

in processing soybeans was calculated based on data from Argonne National Laboratories

(2007), as shown, below, in Table 6.

Energy for Soybean Processing

Soybean oil 1143.7 BTU/lb

Soybean meal 1143.7 BTU/lb

Table 6: Energy used to process soybeans fed to chickens. This energy input corresponds to the

energy used to crush and process the raw soybeans and turn them into soybean meal and oil.

The converted energy inputs for farming and soybean processing are shown

below, Table 7, for both the 1998 NREL study and the 2006 GREET study.

Abbreviation

E1f1: Energy in farming

E2f2: Energy in Processing Feedstocks

E3f3: Energy in Transesterification Process

Eb: Energy in biodiesel

NER: Net Energy Ratio

Table 7: Comparison of Energy in Soybeans fed to Chickens

4- Discussion/Conclusions

Based on the findings, there are numerous ways for Piedmont Biofuels to improve

their energy balance ratio. From the production of biodiesel, Piedmont Biofuels generates

various types of wastes. Two that constitutes for the bulk of waste produced are crude

glycerin and wash water. The Sidestreams Group looked at alternative techniques for

dealing with these wastes. Possibly the best option for disposing of the glycerin is to burn

it in a CHP diesel engine modified to burn the glycerin. The CHP engine will save

Piedmont Biofuels money by cutting their energy costs, which will also help improve

their net energy ratio, thus their energy balance. Also, the Sidestreams Group explored

the option of using an anaerobic digester.  Since glycerin and the wash water are good

sources of carbon, they could both be utilized by anaerobic bacteria to produce methane.

Theoretically the methane could be converted into methanol and then used back into the

biodiesel production process.

Another option would be to use the biogas generated by an anaerobic fermenter as

heat energy throughout the production process and sell the compost to local farmers.

This option has the potential to save Piedmont Biofuels money on storage cost and on the

disposal of the two sidestreams.  Both elimination options are still in the experimental

stage of development.  Currently, there are no amounts determined on methanol or biogas

yields from crude glycerin and wash water.  Also, the anaerobic organism used in the

process will impact the yields.  Research will have to be conducted to determine a

suitable organism in order to reach optimal yields.  A feasibility analysis will also have to

be completed to determine if the disposal cost savings and profits made from selling the

compost will out-weigh the cost of setting up the anaerobic digester in the long run.

While these waste elimination options are excellent suggestions that should be explored

further, currently there is no supporting evidence that adopting these practices will have a

positive effect on the energy balance.

For the biodiesel production process, excess methanol is required to force the

reaction to completion, generally 20-100% more methanol than actually required for

reaction. Piedmont Biofuels currently uses more than 100% methanol, half of which goes

into the wash water. A study to access the optimal amount of excess methanol that is

required for the reaction may help reduce the energy input of methanol. This would not

only reduce the energy input from methanol, but also energy consumed in recovering

methanol from waste water and glycerol.

Our study revealed that the feedstock, chicken fat, is the major contributor of

energy input. Using different feedstock may be helpful in reducing the energy input and

the net energy ratio in the long run. A study of different feedstocks and their energy

content would be helpful to see which source of feedstock would be optimal for Piedmont

Biofuels. While exploring different feedstock may help bringing up the energy balance,

its availability and cost could be a possible constraint.

Farming + processing energy

GREET NREL

8043.7 3901.8 MJ/ha

1300.1 630.7 BTU/lb

After studying the electricity consumption of the plant, we found that most of the

plants total electricity is consumed by reactor heater and wash water tank heater, about

31% and 21%, respectively. Piedmont Biofuels can improve the energy balance by

reducing the electricity consumption in these two heaters, possibly by using more

efficient pumps and motors. Electricity consumption can also be reduced by using

cavitation technology combined with enzymatic biodiesel production, as suggested by the

SBIR group.

Also, transesterification requires a large amount of heat, elevated temperatures,

excess methanol, and high grade feedstock. Low grade feedstock have high free fatty acid

(FFA) content which must be removed, before transesterification, using acid to avoid

soap formation that interfere in the transesterification reaction. The energy balance can be

improved by using biological catalyst, an enzyme, for the biodiesel production. The

Enzymatic Group studied the use of lipase in biodiesel production. The group suggested

that using lipase would reduce the amount of heat and electricity required for the

transesterification process. Also, the use of low grade feedstock would not pose a

problem because of absence of alkali in the reaction and hence, no soap formation. This

would indeed increase the yield as enzyme can convert FFAs to methylesters without

producing soap. Also, enzymes do not require water removal from feedstock which

otherwise leads to the production of soap. Using enzymes would reduce the amount of

methanol needed for the process and would allow the use of low-grade cost effective

feedstock. Enzymatic biodiesel production has several advantages but the cost and

feasibility of using enzymes would need to be determined for use in an industrial scale.

Also, use of enzymes is still in the research phase and there is no evidence of enzymatic

biodiesel production in commercial scale.

In conclusion, Piedmont Biofuels 2009 process clearly indicates an improved

production process in terms of cost and energy due to implementation of biorefinery

model. However, Piedmont Biofuels could improve the biodiesel production by exploring

other feedstocks, improving electricity and methanol consumption, and considering

burning the glycerin wash product in a CHP engine. Even though use of glycerol and

wash water for production of methanol or biogas is still under research, this option could

prove to be very effective in improving Piedmont Biofuel’s energy balance. Finally, the

use of enzymes may seem optimistic but enzymatic biodiesel production could improve

the process, making it more energy favorable and less harsh to the environment.

5- References

Andress, David. (2002). Ethanol Energy Balances. UT-Battelle LLC and

Office of Biomass Programs Energy Efficiency and Renewable Energy

U.S. Department of Energy. Subcontract 4000006704 <

http://www.oregon.gov/ENERGY/RENEW/Biomass/docs/FORUM/EthanolEnerg

yBalance.pdf>.

Argonne National Laboratory. 2006. The greenhouse gases, regulated emissions, and

energy use in transportation (GREET) model. Version 1.6. Argonne, Ill.: Argonne

National Laboratory. Available at: www.transportation.anl.gov/software/GREET/

da Costa, R. S., Lora, E. E. S. (2006). The energy balance in the production of palm oil

biodiesel – two case studies: Brazil and Colombia. Federal University of

Itajubá/Excellence Group in Thermal and Distributed Generation NEST

(IEM/UNIFEI), Oil Palm Research Center CENIPALMA/Colombia and Bahia

Federal University  UFBA. < http://www.svebio.se/attachments/33/295.pdf>.

Energy Input Labeling Standards. (February 23, 2008).

< http://www.technocracy.org/EIL_20082_v1_2.pdf >.

Petitte, J. Personal interview. 9 March 2009.

Pradhan, A., Shrestha, D. S., Van Gerpen, J., Duffield, J. (2007). The Energy Balance

of Soybean Oil Biodiesel Production: A Review of Past Studies. Food & Process

Engineering Institute Division of ASABE. Transactions of the ASABE. 51(1):

185-194.

Sheehan, J., V. Camobreco, J. Duffield, M. Graboski, and H. Shapouri. (1998). Life cycle

inventory of biodiesel and petroleum diesel for use in an urban bus. NREL/SR-

580-24089 Golden, CO: National Renewable Energy Laboratory. U.S.

Department of Energy.

Wang et al., Life-Cycle Assessment of Energy and Greenhouse Gas Effects of Soybean-

Derived Biodiesel and Renewable Fuels, ANL, 2008

Acknowledgments: The authors would like to thank the following people in their guidance

for the production of this research: From Piedmont Biofuels: Rachel Burton, Lyle Estill,

Amanda Egdorf-Sand, Leif Forer, Russell Harper, Greg Austic, from Southern Energy

Management: Scott Hoover; and finally from NC State University: Dr.Steven D Terry

(Mechanical and Aerospace Engineering Department), Dr. Alex Hobbs (NCSU Solar

House), & Dr. Michael Flickinger (BTEC).

Comments are closed.