Design and Techno-Economic Analysis of A Process For Transforming Pig Manure Into
Design and Techno-Economic Analysis of A Process For Transforming Pig Manure Into
This study involves the design and techno-economic analysis of a process that couples biodrying and conventional drying to increase the
economic value of liquid pig manure sludge, by transforming it into a stable organic fertilizer. The process has significant environmental benefits
compared to other disposal methods, especially considering recent legislative changes restricting the spreading of liquid pig manure, and at the
same time provides a return on investment. It was found that the rate of return for the process was 14%, but could reach 25% if the price of
the fertilizer product increases as expected.
On a effectué la conception et l’analyse technico-économique d’un procédé qui couple le bioséchage et le séchage conventionnel pour accroî-
tre la valeur économique de la boue de lisier de porc liquide, en le transformant en un fertilisant organique stable. Ce procédé présente des
avantages environnementaux significatifs comparativement à d’autres méthodes d’élimination, surtout si on considère la nouvelle législation qui
restreint l’épandage du lisier de porc, d’autant qu’il s’avère rentable. On a trouvé que le retour sur investissement était de 14%, mais qu’il
pourrait atteindre 25% si le prix du fertilisant augmente comme prévu.
INTRODUCTION liquid manure have curbed the growth of the pork industry. In
Quebec for example, legislation has progressed rapidly to deal
F
or a number of years, the pork industry has faced
with these environmental pressures. Each year, there are approxi-
challenges related to the treatment and disposal of liquid
mately 6.6 million cubic metres of pig manure produced in
pig manure. The public is increasingly less tolerant of
Quebec (BAPE, 2003). Studies conducted in the regions of
environmental problems that result from pig farming. For
Chaudière-Appalaches, Assomption and the Yamaska River of
example, odours generated by the industry hinder the establish-
Quebec estimate that the quantity of farm fertilizer produced is in
ment of new pig farms near residential areas—a “not in my
excess of the agronomic demand for cultivation (Couture et al.,
backyard” phenomenon. Moreover, environmental regulations
1992). There is clearly a pressing need to find alternative disposal
are becoming increasingly strict, representing a risk factor for
strategies, and in particular innovative solutions that increase the
farmers using traditional spreading methods to dispose of their
economic value of liquid pig manure.
liquid pig manure. In several countries, these social and environ-
Several processes to create a value-added product from liquid
mental problems have impeded the growth of the pork industry.
pig manure have been developed and implemented at pig
Agro-environmental and economic measures including subsidies,
farming operations (Melse and Verdoes, 2005). These processes
taxes and dues, environmental cross compliance, and eco-
generally involve separation methods and/or biochemical
labelling, are becoming increasingly important in the agricul-
reactions such as composting, mechanical or chemical separa-
tural policies of many industrialized countries including Canada,
tion, aerobic or anaerobic biological treatment, and thermal
the United States and several European nations (Debailleul and
treatment. A comparative study of these different treatment
Boutin, 2004; OECD, 2003a).
techniques indicated that economics, as opposed to technology,
The intensification of agriculture can result in water contami-
nation, soil degradation, and the deterioration of habitats and
biodiversity (Boutin, 2004; MENV, 2003). In certain regions where * Author to whom correspondence may be addressed.
pig farms are concentrated, restrictions placed on the spreading of E-mail address: [email protected]
360 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 85, JUNE 2007
was the critical factor determining their success for implementa- • Direct and indirect capital costs were estimated based
tion. All the technologies considered met the minimum require- primarily upon information contained on the Piping and
ments for treatment, whereas their profitability depended on Instrumentation Diagrams (P&ID) and general layouts, as
local economic conditions such as market prospects and disposal well as a careful definition of the battery limits for the project
costs for manure and its products (Melse and Verdoes, 2005). In including the requirements for process, civil, mechanical,
this context, it was among the design project objectives to piping, instrumentation and electrical engineering aspects;
develop a systematic methodology for a risk-based techno- • Operating costs and revenues were estimated based primarily
economic analysis, for designing a process to increase the on the table of unit costs and heat and mass balance quantities
economic value of liquid pig manure. estimated in the process flow diagram (PFD);
The methodology employed in this design study is somewhat • The capital and operating costs were combined to calculate a
similar to those described by Filion et al. (2001) and Khataniar return-on-investment (ROI);
et al. (2002), who used a techno-economic analysis to evaluate • Controllable risk factors were identified related primarily to
the design of processes for the incineration of hazardous wastes the process technologies, for example the process performance
with a plasma burner and the production of gas from hydrates. metrics that were assumed for major equipment items, and as
The design methodology described in this paper is more system- appropriate contingency costs were allowed for and the ROI
atic and it employs design tools to evaluate the critical control- re-evaluated using scenario analysis;
lable and uncontrollable risks. It allows a better focus on the • Uncontrollable risk factors were identified related primarily
solutions to reduce controllable risks and it helps to apprehend to unit costs such as fossil fuel and the cost of capital, and the
uncontrollable risks. This analysis can facilitate the estimation project ROI was re-evaluated using a sensitivity analysis by
of the minimum acceptable rate of return that investors should changing values one variable at a time across a likely range.
consider (Peters and Timmerhaus, 2003). However, the design This methodology involves a careful understanding of the
methodology does not include a lifecycle cost analysis (Geitner basic principles of operation for major equipment items, in order
and Galster, 2000) to evaluate certain equipment operating costs to consider a reasonably low investment cost, while still account-
such as maintenance over the longer term. ing for possible lower operating performance. Reasonable
The design methodology, the cost estimates and the analysis of forecast information must be available for the uncontrollable
the risks pertaining to the base case are initially presented in this risk factors. Several methods have been developed to consider a
paper. The impact of three significant risk factors are subsequently number of risks simultaneously, such as the Monte Carlo
examined, namely energy costs, the efficiency of the biodrying simulation (Janssen et al., 2006), however, this was not done for
process and the market value of the fertilizer produced. this study.
VOLUME 85, JUNE 2007 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 361
clarified liquid is sent to a nitrification-de-nitrification reactor insulation to maintain it. The biological heat provided by aerobic
(NDR), where fresh air is injected by high-pressure blowers, and mesophilic and thermophilic micro-organisms naturally present
the carbon is stabilized by bacteria. The temperature in this in biological sludge is essential for the biodrying process. There
reactor must be carefully controlled so as not to exceed 40ºC, in are currently several biodrying process technologies under
order to prevent a reduction in biotreatment efficiency. Above development and installed (Navaee-Ardeh et al., 2006).
40ºC, the growth rate of the mesophilic organisms decreases However, the actual performance of the biodryer is difficult to
quickly (Bailey and Ollis, 1986). The biotreated liquid effluent predict since few comparable units are available on today’s
from this step is fed to a rotating drum operated under vacuum, market. The biodryer was considered to induce technology risk
from which a secondary sludge containing 18% solids is given its uncertain performance. For this study, an analysis
extracted. The two sludges are sent to the second phase of the relating the residence time in the biodryer to the rate of return
process. The liquid and the gas produced in the first train are on investment is presented. The risk associated with the
sent to a tertiary treatment unit and an offgas treatment unit, uncertain overall drying efficiency of the process may be
respectively, for purification before being released to the environ- addressed by oversizing the dryer; thus the incrementally larger
ment. Since installations already exist for this first train to treat investment required for the dryer represents a strategy to account
the liquid manure, the associated risks for this part of the for the technological risk associated with the biodryer.
process have already been well defined and minimized. The process described earlier can supply some or all of the
The sludge drying train involves the processing of the mixed heat required to dry the mixed sludge to targeted levels. The
sludge into a marketable product. The primary and secondary system components included in the techno-economic analysis
sludges are mixed together before being fed to a biodryer and were the mixer, biodryer and dryer, as well as auxiliary
then a dryer. A recirculation loop from the dryer outlet to the equipment especially for energy management (not shown on the
entrance of the mixture ahead of the biodryer is used to obtain process schematic) such as the heat pump, heat exchangers,
a high porosity blend with 40% dryness, thus maximizing the burner and cooling system.
efficiency of the biodryer due to better pneumatic conditions and
better contact between the solids and air. The sludge mixture
stays in the biodryer for a period of about 7 d, allowing for PROCESS OPTIONS CONSIDERED
homogeneous biological stabilization of the product, and Two process options were selected after a team brainstorming
boosting dryness levels to 50%. Following this step, the dryer and a preliminary analysis of many options regarding energy
further elevates the dryness levels to about 87%. recovery and feasibility. Following selection, Option A and
In the biodrying process, there are two main mechanisms by Option B were analyzed with regard to different risk factors.
which moisture is removed: convection and diffusion. Convection In Option A shown in Figure 3A, controllable risks are
is mainly related to the air flow rate, temperature and humidity, minimized while energy is recovered. Warm air from the NDR
whereas diffusion is related mostly to the moisture distribution (T = 37ºC) is combined with warm air from the dryer
within the particle. In the process, a porous matrix is dried by (T = 55ºC) in order to preheat the fresh air make-up stream
forced air convection. When the surface water concentration being sent to tertiary treatment. After exiting the heat exchanger,
becomes less than that of bound moisture, a water gradient is this air stream is heated further in a heat pump using the warm
present across the matrix and moisture removal depends on liquid from the bioreactor in a closed-loop system. This measure
capillary and diffusion forces. The biodrying reactor is self- also helps to control the internal temperature of the bioreactor,
heating, i.e., it relies on microbial heat production to obtain the which is essential to ensure biological performance. In order to
required process temperatures, and relies only on external minimize the process risks related to energy shortages, a fuel oil
Figure 2. Simplified process flow diagram of the liquid pig manure treatment process
362 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 85, JUNE 2007
Figure 3A. Process flow diagram that minimizes process risks
VOLUME 85, JUNE 2007 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 363
burner was added before the dryer as a precaution. This The quality of the finished product was assumed to be equiva-
additional step may be required periodically, especially during lent for each option. The figures and tables presented show costs
cold temperatures, to raise the temperature of the air to 65ºC in 2005 Canadian dollars. This is considered to be a feasibility-
before it enters the dryer. grade design study whose cost estimates are accurate to ± 30%.
Option B minimizes investment costs as presented in Figure As mentioned earlier, the costs considered in this study
3B. This option employs two heat exchangers to recover most of include those related to the sludge drying train and energy
the available energy from the process. Thus, the heat pump is no recovery equipment. Certain costs, such as those related to
longer required for the NDR, whose cooling is accomplished building expenditures and working capital, were assumed to
with a water cooling system (not shown in the figure). This have been accounted for in the liquid manure treatment train.
feature significantly reduces the investment costs. The first heat Moreover, the product supply chain costs are not taken into
exchanger uses air from the bioreactor (T = 37ºC), the biodryer account because the final product is purchased on-site by
(T = 60ºC), and the dryer (T = 55ºC) to preheat the fresh air fertilizer wholesalers.
stream. The air stream is heated further in a second heat
exchanger using air from the air compressors (T = 82ºC), before
being fed to the dryer at a final temperature of 55ºC. The fuel oil
CAPITAL EQUIPMENT
burner is also included in this option. The costs of the biodryer, dryer, mixer, burner, conveyor system,
cooling system, heat exchangers and heat pump were estimated
based on quotations provided by equipment suppliers, evalua-
PRELIMINARY DESIGN RESULTS tions by experts and heuristical cost relationships. The cost
Various process simulations were carried out using HYSYS difference between Options A and B is due primarily to the fact
(Aspen Technology Inc., 1998) in order to assess in more detail that energy from the bioreactor in Option B is not recovered with
the availability of energy throughout the year. a heat pump, but rather with an additional heat exchanger
A fresh liquid manure feed of 50 t/d was assumed to produce paired with a cooling system.
1.6 t/d of organic base. It was also assumed that the optimal The cost of the biodryer was determined using the “six-tenths
drying temperature for the organic base was in the range of 55 factor rule,” an estimation method for equipment costs based on
to 60ºC, and that the amount of energy required was equivalent scaling (Peters and Timmerhaus, 2003). The cost of an industrial
to that necessary to evaporate enough water to raise the dryness biodryer (including control panel), capable of treating different
level from 50% to 87%, the specification for rendering it a types of organic material and with a capacity of 600 t/d, was
marketable product. According to these design constraints, the assumed. In order to allow for the possibility of treating greater
dryer required 1500 kWh/d. quantities of liquid manure during periods of maintenance or
With the goal of quantifying temperature variations of the air retrofitting, a capacity safety factor of 15% was considered,
stream entering the dryer as a function of the outside tempera- along with a residence time of 7 d.
ture, a meteorological database for a geographical area located The cost of the dryer was estimated based on discussions with
close to the existing facilities was used. several suppliers concerning the treatment capacity, the efficiency,
Based on the average meteorological conditions over the past and the desired degree of automation. For the air-to-air shell-
30 years, the temperature of the heated fresh air stream entering and-tube heat exchanger, a cost estimate was obtained from the
the dryer in Option A varies between 50ºC in January to 77ºC Enerquin. The characteristics of the required heat pump differ
in July. Supplementary energy in the range of 55kW to 120kW from those of commercial models, resulting in a significant
is therefore needed by the burner. Under the same conditions, technological risk for Option A. The cost estimate of the heat
the temperature of the air stream entering the dryer in Option B pump provided is an order-of-magnitude value since it was
varies between 60ºC and 63ºC. The burner was also included in difficult to evaluate the development costs.
Option B in order to minimize controllable risks. However, if it The costs for several pieces of supplementary equipment were
is found that the use of the burner is not needed for the operation also considered. The cost of the burner was determined using a
of the first full-scale facility, it is possible that this feature may relationship between the purchase cost and the capacity of the
not be included in future installations. burner (Peters and Timmerhaus, 2003). The costs associated
with the conveyor system were estimated based on empirical
relationships (Peters and Timmerhaus, 2003). The general
COST ESTIMATION layout helped to determine the length (20 feet) and width
Techno-economic analyses can permit the comparison of two (14 feet) required for sludge transport. The cost for the coiled
process options with regard to various risk factors according to cooling system, used in Option B to cool down the NDR when its
the systematic design methodology used in this study. Sensitivity temperature exceeds 37ºC, was evaluated by a supplier from
analyses are performed in order to identify the parameters that Gestion Air Inc.
have the greatest influence on indicators of economic perform-
ance. The performance indicator selected to analyze the profita-
bility of the project was the simple Return on Investment (ROI) CAPITAL COSTS
(Peters and Timmerhaus, 2003) without taking taxes and Although the investment costs for both options are of the same
depreciation into account, since the project being analyzed is an order of magnitude, Option B is less expensive than Option A.
add-on to a global process for liquid pig manure treatment.
Internal Rate of Return (IRR) and other return criteria might also
be considered, depending on company preferences.
UTILITY REQUIREMENTS
The recovery of energy from the liquid manure treatment train
Gross annual revenue − operating costs
ROI = (1) allows for a significant reduction in the utilities requirement for
total capital cost the sludge drying train. These utilities may be divided into two
364 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 85, JUNE 2007
Table 1. Summary of capital cost estimate Table 3. Summary of projects risks considered
Total indirect cost 0.10 (PE) 233 000 213 000 monitor the day-to-day performance of the process, a remote
management system has been integrated into the process. The
Total capital cost depreciation is calculated linearly without residual value over 20
1.21 (PE) $1 268 000 $1 159 000 years, the period of time corresponding to the lifespan of
PE: Purchased equipment the equipment.
I: Included in the liquid manure treatment train
REVENUES
Table 2. Summary of operating cost estimate The sale of the dry sludge finished product price has been
estimated at $200/t. A more conservative estimate, based only on
Unit cost Option A Option B the content of nutritional elements, yields a price of $160/t.
Utilities However, these values do not consider the potential for environ-
Electricity $0.070/kWh $5 900 $0 mental certification, and the added value of such a product, which
Fuel oil $0.076/kWh 1 700 1 700 could potentially boost the sale price to $300/t according to
Maintenance, 25 830 25 800 fertilizer market experts. The price of the organic base can thus
repairs, and insurance vary depending on market demand and on the sludge characteris-
Remote management, 17 200 17 200 tics, which represents an important uncontrollable risk.
site visits, MENV fees, analyses A second potential source of revenue is in the form of CO2
Depreciation (5% of the 71 100 64 900 credits offered by the Government of Canada for a mitigation of
total capital cost) greenhouse gas emissions. The liquid manure treatment process
and the sludge drying train both reduce CO2 emissions compared
different categories: those required on a daily basis, which to the spreading of liquid manure on agricultural lands.
depends on production levels, and those needed periodically The cost estimates shown in the tables and the estimated
when weather conditions hinder attaining the desired tempera- revenues were used to calculate the ROI for both options, yielding
ture for drying. values of 4.9% and 5.7% for Options A and B, respectively.
Electricity is required for elements specific to Option A, such
as the heat pump that operates year-round. Furthermore, both
options need a source of energy in case of an emergency. A fuel PROJECT RISK ASSESSMENT
oil burner has been assumed that is operated in the winter when Project risk analysis is critical given the innovative nature of the
the energy recovered from the process is insufficient to ensure process and the influence of the design parameters on its profit-
adequate drying. There is a good deal of uncertainty with respect ability. Controllable and uncontrollable risks considered in this
to the necessary quantity of supplemental energy required and assessment are listed in Table 3 and were determined using the
consequently, the utility costs. In Option B, cooling water is also systematic design methodology presented before.
used as a utility. However, the costs associated with this are low, Amongst these risks, three are associated with important
since the NDR only requires cooling when temperatures are aspects of the process (namely operating costs, revenues and
especially high. Thus, the cost for cooling water was not major equipment costs) and are further analyzed here.
included in the operating costs. After the implementation of the
first sludge drying train, the quantity of cooling water required
can be determined more precisely.
COST OF ELECTRICITY
A significant increase in the cost of electricity is being forecast
in Quebec due to increases in the demand for energy and rising
OPERATING COSTS production costs. Several annual cost increase scenarios were
Operating costs include raw materials, labour, depreciation, examined, as shown in Figure 4. The rate of return on invest-
maintenance costs and utilities. Instead of using an operator to ment was calculated over a 20-year period with constant annual
VOLUME 85, JUNE 2007 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 365
Figure 4. Effect of the annual increase in the cost of electricity over
20 years on the rate of return on investment for Option A
Figure 6. Rate of return on investment as a function of the residence
time in the biodryer
366 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 85, JUNE 2007
capacity equipment units. In the case of the biodryer, the
principle of mechanical operation was simplified in order to
facilitate the automation and control features. These design
changes were identified as a consequence of the systematic
design methodology employed.
Based on the calculation methods presented in the cost
estimate section, the new total investment costs were reduced to
$430 000 and $350 000 for Options A and B, respectively. The
corresponding rates of return on investment thus rose to 10.5
and 14%, respectively. Figures 7 and 8 show that if the sale price
were to attain a value of $300/t for the fertilizer, the ROI would
reach 25%.
CONCLUSIONS
A systematic design methodology was applied to examine the
capital and operating costs associated with a novel biodrying-
drying process, for raising the solids context of mixed sludge from
liquid pig manure treatment to 87% where the sludge can be sold
as an added-value product. Two process options were considered
(A and B), which can be differentiated according to energy
systems and purchased equipment. It was found that the total
capital cost of Option A was $1 268 000 and of Option B was
$1 159 000, providing ROIs of 4.9% and 5.7%, respectively.
A systematic design methodology is presented which explores
Figure 7. Reduced capital cost for biodryer and dryer: ROI as a function key parameters that may have a significant influence on project
of the sale price of the dried organic base ROI. It was found that the cost of energy has a negligible impact
on ROI, which greatly reduces the uncontrollable risks.
Furthermore, it is not worthwhile to recover low-quality energy
within the process (Option A). The results of a process risk
analysis indicate that by optimizing the dryer and biodryer, the
rate of return on investment can be doubled. Furthermore, the
risks associated with the purchase of these major equipment
items are limited. As indicated by the analysis of the variation in
the efficiency of the biodryer, even if the residence time in the
new pieces of equipment were doubled, they would still remain
more profitable compared to the standard units. Lowering the
costs of the biodryer and dryer contributed greatly to a reduction
in the investment cost, which caused the ROI to increase from
5.7% to 14%. If the fertilizer price rises to $300/t, the ROI would
reach 25%. In summary, Option B is recommended due to the
incrementally more attractive ROI compared to Option A, better
recovery of the process energy (which reduced the temperature
variation at the dryer intake caused by metrological conditions)
and lower technological risks (related to Option A heat pump).
The calculations presented do not take into account important
beneficial effects of the treatment process, including mitigation
of environmental risks, elimination of social nuisances such as
odour, improvement of the corporate image of pork producers,
Figure 8. Reduced capital cost for biodryer and dryer: ROI as a function stabilization of the environmental management costs compared
of residence time in the biodryer to the possibility of uncontrolled inflation of costs for current
methods, etc.
tural process equipment), a selection of less expensive materi-
als of construction (galvanized steel versus stainless-steel for
parts that are not directly in contact with the manure, plastic REFERENCES
versus stainless-steel for part requiring low resistance but in Bailey J. E. and D. F. Ollis, “Biochemical Engineering
contact with the manure—and assuming no increase in mainte- Fundamentals,” 2nd ed., McGraw-Hill, New York, NY
nance costs) and reduced system controls. The price of the new (1986), pp. 392–393.
equipment can be negotiated directly with the manufacturers Boutin, D., “Réconcilier le soutien à l’agriculture et la protec-
as opposed to the price of the industrial process equipment that tion de l’environnement – Tendances et perspectives,” in
involves intermediaries. Furthermore, the capacity of the “67e Congrès de l’Ordre des agronomes du Québec, Vers une
equipment was targeted to more closely meet the specific politique agricole visionnaire,” Sherbrooke, QC, June 11,
design needs, as opposed to the initial design involving larger 2004, Ministère de l’Environnement du Québec (2004).
VOLUME 85, JUNE 2007 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 367
Bureau d’audiences publiques sur l’environnement (BAPE),
“L’état de la situation de la production porcine au Québec,”
Annexe FULI49 rapport d’enquête et d’audience publique,
no. 179, Québec government, 104 (2003).
Couture, J.-N., M. Trudelle and M. Perron, “Situation de la
valorisation des engrais de ferme,” in “Compte-rendu des
conférences présentées au Colloque sur la gestion des
fumiers,” Conseil des productions végétales du Québec,
Drummondville, QC, October 20–21 1992, Québec (1992),
pp. 183–201.
Debailleul, G. and D. Boutin, “La sévérité de la réglementation
environnementale québécoise dans le domaine des produc-
tions animales : mythe ou réalité?,” Vecteur Environnement
37(2), 31–35 (2004).
Filion, J., R. J. Munz and E. D. Salin, “A Feasibility Study
of Adaptive Plasma-Assisted Incineration– Part 2: Techno-
Economic Potential Assessment,” Can. J. Chem. Eng. 79(4),
618–625 (2001).
Geitner, F. and D. Galster, “Using Life-Cycle Costing Tools,”
Chem. Eng. 107(2), 80–86 (2000).
Janssen, M., F. Cornejo, K. Riemer, H.-C. Lavallée and P. R.
Stuart, “Techno-Economic Considerations for Dip Production
Increase and Implementation of Cogenenration at an
Integrated Newsprint Mill,” Pulp Paper Canada, 107(7) 33-37
(2006).
Khataniar, S., V. A. Kamath, S. D. Omenihu, S. L. Patil and
A. Y. Dandekar, “Modelling and Economic Analysis of Gas
Production from Hydrates by Depressurization Method,”
Can. J. Chem. Eng. 80(1), 135–143 (2002).
Melse, R. W. and N. Verdoes, “Evaluation of Four Farm-Scale
Systems for the Treatment of Liquid Pig Manure,” Biosystems
Eng. 92(1), 47–57 (2005).
Ministère de l’Environnement, “Synthèse des informations
environnementales disponibles en matière agricole au
Québec,” Direction des politiques du secteur agricole,
ministère de l’Environnement, Québec, Envirodoq
ENV/2003/0025 (2003), p. 143.
Navaee-Ardeh, S., F. Bertrand and P. R. Stuart, “Emerging
Biodrying Technology for the Drying of Pulp and Paper
Mixed Sludges,” in Drying Technology, 24(7), 863–878
(2006)
Organisation for Economic Co-operation and Development
(OECD), “Agri-Environmental Policy Measures: Overview of
Developments,” COM/AGR/CA/ENV/EPOC(2002)95/FINAL
(2003), pp. 5–6.
Peters, M. S. and K. D. Timmerhaus, “Plant Design and
Economics for Chemical Engineers,” 5th ed., McGraw-Hill,
New York, NY (2003).
368 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 85, JUNE 2007