You are on page 1of 21

See

discussions, stats, and author profiles for this publication at:


https://www.researchgate.net/publication/223847703

Biomass power cost and optimum plant


size in western Canada

Article in Biomass and Bioenergy June 2003


DOI: 10.1016/S0961-9534(02)00149-6

CITATIONS READS

184 274

3 authors, including:

Amit Kumar Peter C Flynn


National Institute of Technology, Kuruk University of Alberta
81 PUBLICATIONS 1,794 CITATIONS 59 PUBLICATIONS 1,882 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

The Material Properties of Tick Exoskeleton (which is unrelated to my previous work on


renewable energy economics) View project

All content following this page was uploaded by Peter C Flynn on 27 March 2017.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
Available online at www.sciencedirect.com

Biomass and Bioenergy 24 (2003) 445 464

Biomass power cost and optimum plant size in western Canada


Amit Kumar, Jay B. Cameron, Peter C. Flynn;1
Department of Mechanical Engineering, University of Alberta, Edmonton, AB, Canada T6G 2G8
Received 5 December 2001; received in revised form 21 October 2002; accepted 21 October 2002

Abstract

The power cost and optimum plant size for power plants using three biomass fuels in western Canada were determined.
The three fuels are biomass from agricultural residues (grain straw), whole boreal forest, and forest harvest residues from
existing lumber and pulp operations (limbs and tops). Forest harvest residues have the smallest economic size, 137 MW, and
the highest power cost, $63:00 MWh1 (Year 2000 US$). The optimum size for agricultural residues is 450 MW (the largest
single biomass unit judged feasible in this study), and the power cost is $50:30 MWh1 . If a larger biomass boiler could be
built, the optimum project size for straw would be 628 MW. Whole forest harvesting has an optimum size of 900 MW (two
maximum sized units), and a power cost of $47:16 MWh1 without nutrient replacement. However, power cost versus size
from whole forest is essentially <at from 450 MW ($47:76 MWh1 ) to 3150 MW ($48:86 MWh1 ), so the optimum size
is better thought of as a wide range.
None of these projects are economic today, but could become so with a greenhouse gas credit. All biomass cases show
some <atness in the pro>le of power cost vs. plant capacity. This occurs because the reduction in capital cost per unit
capacity with increasing capacity is o?set by increasing biomass transportation cost as the area from which biomass is drawn
increases. This in turn means that smaller than optimum plants can be built with only a minor cost penalty. Both the yield
of biomass per unit area and the location of the biomass have an impact on power cost and optimum size. Agricultural and
forest harvest residues are transported over existing road networks, whereas the whole forest harvest requires new roads and
has a location remote from existing transmission lines. Nutrient replacement in the whole forest case would make power
from the forest comparable in cost to power from straw.
? 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Biomass usage; Agricultural residues; Forest harvest residues; Wood waste; Power generation; Carbon credit; Optimum size;
Logistics

1. Introduction of greenhouse gas (GHG) emissions from fossil


fuel utilization. In some locations, including western
Biomass usage, speci>cally capturing energy from Canada, good data on the cost of using biomass is
biomass that would otherwise decay, is one of many not available, and this leads to a high degree of un-
options available to mitigate the impact of the buildup certainty in the cost of GHG credits that would be
required to support such a facility. Western Canada,
Corresponding
in particular the Province of Alberta, is a particu-
author. Tel.: +1-780-492-6438; fax:
+1-780-492-2200.
larly relevant place to evaluate the economics of
E-mail address: peter.<ynn@ualberta.ca (P.C. Flynn). generating power from biomass for three reasons.
1 Poole Chair in Management for Engineers. First, Alberta has a growing power demand and is an

0961-9534/03/$ - see front matter ? 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 9 6 1 - 9 5 3 4 ( 0 2 ) 0 0 1 4 9 - 6
446 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

area of active development of new coal based power Our research is based on existing techniques of
plants, with one project of a single 450 MW unit harvesting biomass and commercially proven direct
and a second project of two 450 MW units approved combustion technologies for its utilization. Alter-
and under construction. Second, the region has abun- nate technologies can be used to generate power (for
dant biomass resources, both agricultural and forest. example, gasi>cation vs. direct combustion), and al-
Third, the region has a large oil and gas resource that ternate end uses can be identi>ed (for example the
is being exploited for industrial, domestic and trans- production of liquid fuel byproducts). These tech-
portation fuels, and continued development of this nologies are at the demonstration or research stage.
resource may well depend on developing e?ective If power from biomass were implemented on a large
GHG o?sets. The combination of these three factors scale in western Canada, specialized techniques and
makes western Canada an ideal location for imple- equipment might emerge that would reduce harvest
menting power from biomass at a full commercial costs; examples include simultaneous collection of
scale. straw and grain and simultaneous chipping of for-
The purpose of our research has been to estimate est residues at the time of timber/pulp harvest. The
the cost and evaluate the cost sensitivities for major potential impact of such future improvements is not
biomass utilization projects located in the Province of factored into this study, but will be evaluated in
Alberta. Our research has focused on major biomass future work.
resources located within western Canada that are avail- Power from biomass is not economic today in
able in signi>cant quantities for future power genera- western Canada, where power is generated from a
tion. Speci>cally, three such sources were identi>ed: large base of hydroelectric, gas >red, and base-load
agricultural residues, forest biomass from harvesting mine-mouth coal >red plants. Hence, one key mea-
of the whole forest, and the residues from harvesting sure of cost of biomass is the carbon credit (as
forests for lumber and pulp. Each of these sources is $ tonne1 CO2 abated) required to equalize the
discussed below in more detail. Mill residues from cost of power from a biomass plant with current
processing of lumber and pulp (for example, bark and alternatives. In e?ect, this is the premium asso-
sawdust) were not evaluated, because these are widely ciated with the mitigation of GHG. Because coal
utilized today and recent volatility in the cost of natu- based power projects are under active develop-
ral gas has led to intensive development of additional ment and represent the current marginal power
projects based on this resource. plant fuel of choice, GHG credits are calculated
Our research has focused on a common end use for in comparison to a new coal based power plant
biomass: power generation from direct combustion. using conventional combustion supercritical boiler
Holding end use constant for the three biomass sources technology.
allows an assessment of the relative value of these Previous studies have assessed biomass economics
three biomass resources. It is also key to assessing from the perspective of general models [15]. Dorn-
the comparative economic optimum size of a biomass burg and Faaji have developed a detailed study of
power generation facility, which is fuel speci>c. In small to medium scale biomass plants in a Dutch set-
generating power, biomass fuels di?er from coal: the ting [6]. This study applies the general methodology
variable cost of fuel per unit capacity depends on the to western Canada. Good regional data is available on
capacity of the facility using the biomass, whereas for the cost of harvest and transport of biomass, including
coal >red plants, whether mine mouth (i.e. the power costs of loading and unloading that have not always
plant is located at the coal mine) or remote, the vari- been considered in previous studies. Western Canada
able fuel cost is either negligible or virtually indepen- is also the site of both recent and current major en-
dent of capacity. The variable component of biomass ergy projects, and good data is available on construc-
fuels is related to their transportation cost. Because tion costs for both developed and remote locations.
biomass has a signi>cant variable fuel cost component Hence, this study draws on actual data to determine
that varies with plant size and coal does not, the eco- the cost di?erence for substituting biomass for coal at
nomic optimum size characteristics are quite di?erent an optimum plant size in a region of active coal power
for the two fuels. development.
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 447

2. Biomass sources and nutrient replacement study an average value of 0.80 was used. Hence, for
this study a straw production density of 0:416 tonnes
2.1. Agricultural residues of dry straw per gross hectare in a district was used.
Agricultural residues contain nutrients that re-
The largest concentrated source of >eld-based agri- turn to the soil during the decomposition process.
cultural residues in western Canada is straw from Wheat/barley crops in Alberta are fertilized today, but
crops such as wheat and barley. Note that some barley recovery of residues for usage as fuel would require a
is grown as green feed, i.e. it is harvested whole higher application rate of fertilizer, which is included
and used as silage for animal feed. However, where as a cost in this study. Prairie soils are glacial till,
wheat or barley is threshed, straw is a byproduct that with abundance of calcium and usually of trace min-
is typically laid back on the >eld during the com- erals, so nutrient replacement costs are factored only
bining process. Some of this straw is subsequently for nitrogen, phosphorous, potassium and sulfur.
collected for use as bedding or a feed supplement, Agricultural residues are used today as a fuel for
but most is left to rot on the >eld. Except in cases heat and combined heat and power plants in Europe
of highly sloped soils or unusually high wind areas, [1115]. In some cases these plants are economical be-
straw cover on >elds does not impact erosion con- cause the recovery of straw is required for agricultural
trol if proper stubble height is maintained [7]. Based reasons and is produced in great excess to its needs as
on published yield data [8], the Province of Alberta an animal bedding material or feed supplement. For
alone could support 2000 MW of power generation these cases use of the straw as fuel is a means of waste
from currently uncollected straw. disposal as well as energy recovery, and the straw col-
A key study demonstrated that in black soils in lection costs are a sunk cost not charged to the power
western Canada repetitive straw recovery did not re- generation cost.
duce soil carbon levels [9], presumably because the In this study, agricultural residue cost is based on
contribution of residual stubble and roots to soil car- full replacement of removed nutrients at cost plus full
bon o?sets losses. Hence, the carbon credit available recovery by the farmer of all costs associated with
from recovery of straw in these areas is not partially harvesting straw, including labor and capital recovery
o?set by loss of carbon sequestered in the soil. Agri- for equipment usage. In addition, a market premium
cultural records for six years for three districts in cen- of $4 dry tonne1 is placed on biomass, as discussed
tral Alberta with black soil showed a range of grain below. The resulting value for straw at the >eld used
production per gross hectare from 0.42 to 0:81 tonnes in this study is about 150% of the current value of
[8]. Note that gross hectares include all area within the market value of straw in western Canada. Security of
district, including towns, roads, and land cultivated to fuel supply is a major concern of any developer of a
non-grain uses. For this study an average grain yield power plant, and steps that would help address this
of 0:52 tonnes of grain per gross hectare was used, risk are discussed below.
equivalent to the lower quartile of yields during the
period for which data was recorded. Straw production 2.2. Whole forest biomass
for these areas ranged from 0.75 to 1.05 units of dry
straw 2 per unit of as-delivered grain [9,10]; in this The second source of biomass in this study is whole
forest biomass from a dedicated forest plot, with the
2 Note that for all biomass in this study the reported yields or
power plant located centrally within the plot. Note
weights are on a dry weight basis, except as noted, i.e. actual wet
that in the Province of Alberta the majority of forested
yields are adjusted to zero moisture content (the forestry industry areas are owned and controlled by government, and
refers to bone dry wood, which we treat as identical to a dried these in turn have been committed to existing forestry
tonne). Estimated actual moisture content is 16% for straw, 50% operations (pulp and lumber) under long term manage-
for whole forest chips, and 45% for forest harvest residue chips. ment agreements. Hence, in theory there is no avail-
These estimated actual moisture levels were used in calculating
transportation costs and net heat yields from combustion. The heat
able uncommitted forest area that could be speci>cally
of combustion of biomass per dried tonne varies with species; in harvested for biomass. However, the alignment of
this study a blended value for each biomass fuel is used. forest processing plants (e.g. pulp mills and lumber
448 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

operations) and forest reserves is inexact, and it is generation the limbs and tops (that are left at road-
likely that some excess forest capacity will emerge side in pulp or lumber harvesting) are also recovered,
particularly if faster growing hybrid species are re- requiring the transport of chipped material.
planted after harvesting. Nutrients are not restored in most existing Alberta
Northern forests in Alberta are boreal; two types forest operations, most of which are occurring in areas
of sub-region are the most common: mixed hardwood of >rst cutting. Branches and tops are left in the forest
and spruce. The basis of the whole forest biomass case in current harvesting, however the distribution of these
is a medium yield site [16]. Our assumption is that is usually not uniform. About 80% of harvest opera-
good yield site would be reserved for timber and pulp tions in Alberta skid whole trees to roadside, where
operations; however sensitivity cases for both good they are delimbed and topped. The leaves/needles in
and fair yield site are included. Biomass yields from this trimmed material contain a large portion of the nu-
mature stands of medium yield for mixed hardwood trients, especially nitrogen. The limb and top residue
and spruce are 94 and 74 dry tonnes per net hectare is piled by the side of the logging road, and typically
respectively [16], and these have been blended in this burned at the end of the harvest, which results in a
study to an assumed forest biomass yield of 84 dry loss of the nitrogen to atmosphere. Return of phos-
tonnes of biomass per hectare. Large contiguous ar- phorous, potassium and other trace nutrients in the ash
eas of mixed hardwood and spruce are available in is very limited at best, since ash distribution is rare.
Alberta, and could support a large power plant for a Hence, in current forest harvesting operations virtu-
30+ year life without having to harvest or leap over ally all nutrients in the forest biomass are lost, and are
low yield bog areas, hence an aggregated biomass not replaced.
yield based on the two sub-regions is warranted. The Since the current forestry practice in Alberta, based
study is based on selective clear-cut logging through- on >rst cut, is not to replace nutrients, in this study the
out the dedicated forest plot, resulting in a constant base case does not include a provision for nutrient re-
transportation distance to the power plant over the life placement. However, this is evaluated as a sensitivity
of the plant. since it is a key cost factor if included. For the nu-
This study draws on regionally speci>c detailed trient replacement sensitivity, nitrogen, phosphorous
studies of the costs of recovering forest biomass per- and potassium are replaced. Calcium is not replaced
formed by the Canadian Government, by the For- since it is abundant in boreal forest soils in western
est Engineering Research Institute of Canada, from Canada. Ultimately, as long-term forest management
other literature, and from personal discussions with in European countries has demonstrated, nutrient re-
researchers and equipment suppliers [1729]. In addi- placement is necessary regardless of the end use of the
tion to these sources we have built a detailed model forest biomass. First cut operations take advantage of
of chipping costs for both forest cases in this study. the initial bounty of nutrients in the soil, but this is
For the whole forest case, whole trees are cut and eventually depleted with sustained harvesting.
skidded to a 50/48 Morbark chipper, which prepares Costs for construction of logging roads, and silvi-
chips suitable for direct combustion that are loaded culture costs (replanting, plus nutrient application in
into a waiting chip van. The large-scale chipper is as- the sensitivity case) are included for harvesting the
sumed to operate 5000 h per year, and is fed by a ded- whole forest; these are a signi>cant component of
icated grapple. Based on this speci>c case, a whole overall cost.
forest chipping cost of $2.40 dry tonne1 is calculated. Whole forest biomass cost in this study is thus based
This is considerably lower than other reported values on full recovery of all costs associated with harvesting
in the literature, which range from $8.23 to $14.54 and chipping, including capital recovery, but without
dry tonne1 [3037]. The lower value in this study nutrient replacement. As with agricultural residues, an
arises from the large scale of the chipper (100 green additional market premium of $4 dry tonne1 is placed
tonnes h1 ) and the high number of operating hours on the biomass, but note that this market premium
per year compared to chippers in [3037]. In pulp op- is at the low end of the range of royalty payments
erations the transport of whole trees is an alternative, (stumpage fees) realized from the sale of timber
but in the case of using forest biomass for power cutting rights. Hence traditional pulp and lumber
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 449

operations could, in most market conditions, compete ment. As with the whole forest case, the literature re-
for access to the forest biomass. As a region rich in ports a wide range of chipping costs for residues, from
forest and fossil fuel resources that will likely re- $2.00 to 28.78 dry tonne1 [3032,3842]. For for-
quire GHG o?sets, interesting tradeo?s arise: is the est residues a speci>c case using pilers, loaders and
forest worth more as a low royalty fuel supply that high capacity Nicholson WFP3A chippers with a ca-
enables parallel development of high royalty fossil pacity of 48 green tonnes h1 operating at 5000 h per
fuel projects, as compared to its value for pulp and year gives a total cost of $9.42 dry tonne1 to recover
lumber? Future work will explore these tradeo?s. residues left by the sides of logging roads. Note that
Note that for the whole forest case, security of fuel for forest residues the limit to throughput is the abil-
supply is readily addressed by the granting of cutting ity to feed the material into the chipper, and for this
rights, which are controlled in western Canada by the reason a smaller capacity chipper is used.
Provincial Governments that have retained ownership Costs for construction of logging roads and silvicul-
of the forests. ture are not attributed to the cost of power from har-
vest residues, since the roads and silviculture costs are
2.3. Forest harvest residues required regardless of the disposition of the residues.
In most existing forestry operations in Alberta, nu-
Given that forest resources have a value as >ber trients are not replaced, and the nutrients from harvest
in pulp or lumber, an alternative is to recover har- residues end up being concentrated at roadside or dis-
vest residues. In theory, one could harvest brush and persed in the atmosphere and hence are not available
deadfall as well as limbs and tops, but in practice this to fertilize regrowth, as discussed above. Hence, the
would require a major modi>cation of forest harvest- base case for use of harvest residue does not include
ing, since current practice is to cut and skid trees to a cost for nutrient replacement.
roadside where they are delimbed and topped, whereas Forest harvest residue cost in this study is thus based
brush and deadfall is left in place in the forest. Hence on full recovery of all costs associated with harvest-
the basis of this study is the recovery of limbs and tops ing and chipping, including capital recovery, but with-
from the side of logging roads. These residues range out nutrient replacement. An additional market pre-
from 15% to 25% of the total biomass in the forest. In mium of $4 dry tonne1 is placed on the biomass,
lumber based operations there is a growing emphasis which would result in a direct gain by the company
on cut to >t in the >eld, i.e. trimming logs to the that held the timber cutting rights. In theory a govern-
economic length in the >eld so as to avoid transport- ment could require long-term access to forest harvest
ing waste material to the mill. This practice pushes residues without a premium as a condition of grant-
harvest residues to the 25% range, whereas in some ing cutting rights, thereby reducing the cost of forest
pulp operations it is as low as 15%. Twenty percent harvest residues and addressing long-term security of
residue from a good yield site has been used as the supply.
basis of this study, since current lumber and pulp har-
vesting draws from such sites. This is equivalent to
3. Scope and cost
a blended yield of 24.7 dry tonnes of residue per net
harvested hectare [16]. However, the forest in Alberta Note: all currency >gures in this report are ex-
is harvested on a planned average rotation of 80 120 pressed in US$ and are in base year 2000 unless oth-
years, due to poor soil conditions and a northern cli- erwise noted. Conversion between Canadian and US$,
mate. In this study, a rotation of 100 years is assumed, where required, has been done at the rate of $1 US
giving a yield of forest harvest residues of 0.247 dry = $1.52 Cdn. Costs from the literature have been
tonnes of residue per gross hectare. adjusted to the year 2000 using historical in<ation
Residue material is piled (consolidated from small rates; an in<ation rate of 2% is assumed for 2001 and
roadside piles into larger piles), chipped in the >eld beyond.
and transported to the power plant by chip van truck. The scope of this study is a dedicated power gen-
Chipping of branches and tops is less ePcient than eration plant operating for 30 years using biomass.
chipping whole trees, and requires di?erent equip- Fuel properties are shown in Table 1. Cost factors
450 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

Table 1
Fuel properties

Characteristics Straw Whole forest Forest residues

Moisture content (%) 16 50 45


Heating values (MJ dry kg1 , HHV) 18 [54] 20 [55] 20 [55]
Fuel density during transport (dry kg m3 ) 140 [56] 350 [30] 350 [30]
Nutrient content (%)
Nitrogen 0.66 [9] 0.31a
Phosphorus 0.09 [9] 0.05a
Potassium 1.60 [9] 0.15a
Sulfur 0.23 [9]
Ash (%) 4 [58] 1 [58] 3 [49]
a The data given is for white spruce-subalpine >r forest in Canada [57]. It has been generalized for the boreal forest.

are developed for each element of the scope and are The cost of straw harvesting includes an al-
included in detail in Tables 24. Note that for costs lowance for both equipment and labor, i.e.
a?ected by scale factor, the costs are reported for a in addition to the purchase of biomass the
plant capacity of 450 MW. Some cost factors warrant farmer harvesting straw is paid $10 h1 for
further comment: his time spent in gathering the straw on a
separate pass over the >eld.
Field purchase of biomass: A <at charge of $4 For whole forest biomass, we assume that
dry tonne1 of biomass is assumed, regardless contract harvesting rates cover cutting, skid-
of the heating value of the fuel. Since straw has ding, and >eld chipping of whole trees. For
about 90% of the heating value of the chipped forest harvest residues, we assume that limbs
forest biomass on a dry basis, it has a slightly and tops are stacked and chipped. As noted
higher charge for biomass per unit of power gen- above, chipping of limbs and tops is less eP-
erated, but the impact on power cost is not signif- cient than chipping whole trees, and a higher
icant. Note that $4 dry tonne1 is equivalent in chipping cost is factored into the forest
the whole forest case to $3.50 Canadian m3 of harvest residues case.
recovered lumber or pulpwood; as noted above, Nutrient replacement: The agricultural residue
this is comparable to the lower end of exist- case includes a payment to the farmer for re-
ing royalty (stumpage) fees in the Province of placement of the nutrient content of the straw,
Alberta. while nutrient replacement is not included in
Gathering of biomass in the 3eld: Capital costs the base case for forest biomass, as discussed
for harvesting equipment are not estimated in this above. Note that the farmer is already fertiliz-
study but rather treated as an operating cost that ing grain crops, so the nutrient payment is for
includes capital recovery. This is equivalent to incremental fertilizer only and does not include
assuming that the power plant operator contracts an application cost. For whole forest, the cost of
out harvesting. nutrient replacement is assessed as a sensitivity
For agricultural residues, we assume that case, in which the cost of applying the fertilizer is
farmers harvest the residues, deliver them to included.
roadside as large bales and cover these with Transport of biomass to the power plant site:
tarps to limit moisture ingress. The power For agricultural residues, biomass transport is
plant operator would be responsible for ar- over existing publicly maintained roads. For
ranging pickup of the bales (tarps would be forest harvest residues, this is over existing for-
left by roadside for reuse by the farmer). est roads from pulp and lumber operations that
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 451

Table 2
Production and delivery of biomass

Factor Value Source/comments

Straw
Cost of straw recovery excluding nutrient 8.86 The data is made up of two sources: baling [59] and tarping
replacement ($ dry tonne1 ) [60].
Straw loading and unloading cost 4.00 This value is the sum of the loading and unloading costs. It
($ green tonne1 ) is a blended value of costs taken from the literature [59,60]
and quotes from Alberta based custom straw haulers.
Straw transport cost 0.11 This is a blended value of costs taken from the litera-
($ green tonne1 km1 ) ture [59,60] and quotes from Alberta based custom straw
haulers.
Nutrient spreading cost ($) 0 It is assumed that there are no spreading costs for agri-
cultural residues because the additional nutrients lost to
residue removal can be replaced during existing fertilizing
operations.
Wholeforest
Whole forest harvest cost including skidding In the formula V stands for mean merchantable volume of
to roadside ($ m3 ) 0:5963 per stem. Average merchantable volume is assumed to be
0:9177V
Feeling 90% of the gross volume per stem. Skidding distance is
0:9936V 0:3676
Skidding assumed to be 150 m. Value of V is assumed to be 0:26 m3
per stem based on medium yields of hardwood and spruce
in the boreal forest [35].
Chip loading, unloading and transport cost 0:7585 (2:30 + D is the round-trip road distance from the forest to the re-
($ m3 ) 0:0257D) ceiving plant [35]. In this study the cost has been converted
to green metric tonnes.
Road construction and infrastructure cost 0:7585 + VT is the mean merchantable volume per hectare, where
($ m3 ) (379:24=VT ) T is the mean number of merchantable stems per hectare.
Value of VT has been assumed to be 185:4 m3 ha1
for the boreal forest. The construction cost of roads is
$379:24 ha1 represents the tertiary road network used
only during the year of the harvest. Infrastructure cost of
$0:7585 m3 depends on the amount of labor and machine,
and possibly the merchantable volume per hectare [35].
Silviculture cost ($ ha1 ) 151.69 Many Canadian provinces require that silviculture treat-
ments be performed shortly after harvesting, so that cut
areas are returned to a productive state [35].
Nutrient spreading Cost ($ ha1 ) 73.00 [61]
Chipping cost for whole tree ($ dry tonne1 ) 2.40 Based on detailed study of Morbark 50/48 whole tree chip-
per.
Premium above cost of fuel that is paid to 4.00 Note that US$ 4 dry tonne1 of whole forest biomass
owner as an incentive to collect and sell the is approximately Cnd$ 3:50 m3 of recovered lumber or
fuel ($ dry tonne1 ) pulpwood. This is comparable to the lower end of existing
royalty (stumpage) fees in the Province of Alberta.

Forest residues
Chipping cost of forest residues 9.42 The cost of chipping for forest residues includes forwarding
($ dry tonne1 ) and piling.
Chip loading, unloading and transport 0:7585 (2:30 + D is the round-trip road distance from the forest to the re-
cost ($ m3 ) 0:0257D) ceiving plant [35]. In this study the cost has been converted
to green metric tonnes. The transport cost formula for the
chips in the whole forest case and forest residue case is the
same.
452 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

Table 3
Power plant characteristics and costs

Factor Value Source/comments

Plant life (years) 30


Net plant ePciency (LHV) (%) 34 Internal plant use of power is assumed at 10% of
gross [44,49,53].
Plant operating factor:
Year 1 0.70
Year 2 0.80
Year 3 onwards 0.85

Operating staPng excluding maintenance sta? : StaPng levels are derived from the literature
450 MW or below 8 [49,53,62], and discussions with personnel in the
Above 450 MW; for each additional unit 4 power generation industry. For a plant up to
450 MW, operators per shift are fuel receiver (1),
fuel handlers (2), control room (2), ash handling
plant (1), and other power plant tasks (2). For each
additional unit we add one fuel handler, one ash
handler, and two sta? for other power plant tasks.
The assumed staPng is >ve shifts (10; 400 h per
shift position per year), which allows for vacation
coverage and training.

Power generation capital cost ($ kW1 at 450 MW) This is for a 450 MW direct combustion biomass
Straw plants 1300 power plant determined from the literature [43,49],
Wood plants 1184 existing straw plants [1215], and existing wood
plants [44]. Note that this >gure is more than 50%
higher than comparable >gures for coal based power
generation; the source of this discrepancy is not
obvious and the cost for biomass power generation
is considered conservative, i.e. high.
Average annual labor cost including bene>ts ($ h1 )
Operators 27.00
Administration sta? 27.00
Ash disposal cost Hauling distance for the ash is assumed to be 50 km
for the three cases.
Ash hauling cost ($ dry tonne1 km1 ) 0.114 [26]
Ash disposal cost ($ dry tonne1 ha1 ) 15.90 [26]
Amount of ash disposal (dry tonnes ha1 ) 1 [26]

Transmission charge for remote location ($ MWh1 ) 2:16 The transmission charge for the whole forest case
Capital cost 2:08 has been calculated assuming 300 km of dedicated
Operating cost 0:08 lines carrying 900 MW at a total capital cost of $97
million at 10% capital recovery plus an operating
cost of $408,000 excluding line loss. The cost is for
the power plant running at full load at a capacity
factor of 0.85.
Spread of costs during construction (%) Plant startup is at end of year 3 of construction.
Year 1 20
Year 2 35
Year 3 45
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 453

Table 4
General assumptions

Factor Value Source/comments


Scale factor
Total power plant capacity 20 to 0.75 [43,44]
450 MW
Transmission line capital cost 0.49 0.49 is based on >tting a curve to estimates of 300 km
transmission lines through remote boreal forest at vari-
ous capacities. This value is an exponent.
Transmission line operating cost 0.50 0.5 is an exponent for operating costs and is an estimate
based on consultation with the electrical industry.
Cost of an additional equal sized power 0.95 0.95 is based on conversations with Engineering Pro-
plant unit relative to the >rst curement Construction (EPC) contractors. This value is
not an exponent. It states that additional identical power
plant units only cost 95% as much as the >rst unit [47].
Factor to re<ect capital cost impact for 1.10 1.1 is based on discussions with EPC contractors regard-
remote location ing construction of a power plant in a remote location
[46].
Transmission loss for remote location 3% of generated The value has been estimated based on consultation with
power the electrical industry for a base load 300 km line.
Annual maintenance cost 3% of initial capital The value has been assumed based on blending data
cost per year from existing coal->red units and from studies of
biomass power plants [43,49,52].
Labor surcharge for remote location 1.20
Aggregate pre-tax return on investment 10 %
(blend of debt plus equity)

Site recovery and reclamation costs 20% of initial capital The reclamation cost is escalated and is assumed to be
cost in the 30th year of operation.

Nutrient cost ($ kg1 )


Nitrogen 0.62 [9,63] The nutrient costs are given in cost per unit of
P2 O 5 0.41 fertilizer. To determine the cost of nutrient replacement
K2 O 0.22 one must multiply by the amount of nutrient per unit of
Sulfur 0.26 fertilizer. K2 O is 83% potassium. P2 O5 is 44% Phos-
phorus.

generate the residues. For whole forest biomass, Processing of biomass at the plant site: A small
the cost of road building is charged to the project reserve of biomass is stored on site (equivalent
since there is no existing road infrastructure. to about two weeks operation) to sustain the
As noted biomass projects have a transportation power plant when roads are impassible. Straw is
cost that varies with plant capacity. This arises chopped at the plant site.
because the area from which biomass is drawn Combustion of the biomass in a boiler, with use
is proportional to plant capacity, and the haul of the steam solely for power generation: Full
distance is proportional to the square root of area. capital costs are calculated for power generation,
Biomass economics are thus sensitive to biomass and are adjusted for capacity by a scale factor.
yield: higher yields per unit area reduce the Note that co-generation, the use of low-pressure
area required to sustain a given project size. We steam exhausted from turbo generators for heat-
explore this e?ect in a sensitivity. ing, is not considered in this study.
454 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

Scale factor: The base case unit scale fac- isting biomass combustion units, and the
tor used in this study was 0.75, where scale maximum unit size that is acceptable in
factor is an exponent for adjusting the cost relation to the size of the electrical power
of a power generation unit from one ca- market. We note that the three coal >red
pacity to another (i.e. Cost 2 = Cost 1 units being built in the Province of Alberta
(Capacity2 =Capacity1 )Scale factor ). Scale fac- are all sized at 450 MW, although larger
tors for single boiler biomass power plants coal >red units have been built in other
from the literature range from 0.7 to 0.8 locations, e.g. [48]. The assumption of max-
[4345]; similar values are reported for coal imum unit size is critical for two cases in
[46,47]. Actual cost data is available for this study, where the optimum plant size
a number of straw based plants, although is found to be one or more of the maxi-
comparison is diPcult because the plants mum sized units. This is discussed further
use the steam for heat and power, and the below.
relative mix of these varies from plant to Capital cost: Data were drawn from a
plant [1215]. After manipulating the data variety of actual plant costs and litera-
to adjust for scope, the scale factor is es- ture sources, and show a wide variability
timated at 0.8, but this re<ects plants built [1215,49]. Actual data for straw >red
in a variety of locations that are always units, built for a mix of heat and power,
new to that location and that are small appeared after manipulation for scope and
and built as demonstration units. For that size to be about 20% higher than wood
reason, we have assumed that in a mature based biomass units. Values used in this
large scale facility the scale factor would be study are $1300 kW1 for straw, and
lower. Previous studies have shown some $1184 kW1 for wood and forest residues
disagreement on appropriate range of scale at a size of 450 MW; comparable values
factors; Jenkins [1] has explored a wide for new coal->red plants in Alberta are
range, from zero to 1.0, while Dornburg $850 kW1 . We make two notes on these
and Faaji [6] argue for a narrower range. values. First, many biomass plants built
Based on discussions with >rms that have to date have been demonstration units, for
built major energy facilities, we explore which higher capital costs would be ex-
the impact of scale factor in the range of pected than would be realized with a ma-
0.6 0.9 for a single unit up to 450 MW ture technology. Second, boiler/power plant
size. Over 450 MW, a step change in scale costs for straw and wood are 50% and
factor occurs: the cost of an additional 40% higher than comparable capital costs
identical unit is assumed to be 95% [47] for large coal >red boiler/power plants in
of the >rst unit cost, i.e. the cost of build- western Canada (which has low sulfur coal
ing an incremental identical unit saves 5% that does not required sulfur removal from
on the incremental unit only. This is close <ue gas). Several factors contribute to a
to Jenkins assumption that scale factors higher cost for burning biomass, including
approach unity as project sizes get very higher mass <ow rate of solid fuel, lower
large. <ame temperature (and hence larger con-
Maximum unit size: In this study we have vective to radiant ratio in the boiler) and
assumed that the maximum unit size for a a more corrosive ash [50], but these fac-
biomass >red boiler is 450 MWe . For any tors do not readily equate to such a large
capacity over 450 MW, two or more iden- di?erence in cost as compared to coal.
tical sized units are built, e.g. at 500 MW We are not able to justify the large pre-
two units of 250 MW would be built. This mium in capital cost compared to coal, and
assumption re<ects two qualitative factors: hence the biomass capital cost values may
a judgment re comfort in scale up of ex- be conservative (high). Capital cost of the
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 455

boiler and power plant is thus a source of sion line. In Alberta, the likely location of an
uncertainty; we have run sensitivities on agricultural or forest harvest residue power
capital cost to explore the impact of this generation plant is also in a power load
uncertainty. consuming area, so that there would likely
Location: Alberta has a cold winter, but be no transmission penalty assessed. Hence,
also has a workforce and construction in- no net transmission cost is assigned to the
dustry well used to working productively in generation facility.
cold weather. Hence, no capital cost penalty For the whole forest biomass case, the basis
was applied for climactic conditions. How- of the study is a remote forest plot located
ever, in all cases the plants are suPciently 300 km from existing transmission lines,
remote from major population centers that which requires a dedicated transmission line
construction labor would be housed in a to connect to the existing grid. This trans-
camp, and a provision of $13 million was mission line is assumed to have 3% line
provided for the camp and for workforce loss. The cost of the line is recovered as a
transportation costs at a 450 MW capac- transmission charge; at the optimum sized
ity, and adjusted for scale [46]. The whole whole forest biomass plant, the charge to
forest power plant is built in a remote loca- recover the cost of the transmission line
tion away from existing infrastructure, and is $1:52 MWh1 . The scale factor for the
would have additional costs during con- remote transmission line is 0.5 rather than
struction such as access roads, higher freight the 0.75 >gure used for power generation
costs, higher contractor mobilization and equipment; the 0.5 factor is based on actual
demobilization costs, and a longer construc- estimates for transmission lines at vari-
tion sta? cycle (for example, two weeks in ous scales, and re<ects that clearing of the
and one week out rather than the traditional right of way is required regardless of line
>ve day work week). To account for this, capacity.
capital costs are escalated by 10% for this Operating costs: For the agricultural and forest
case [46]. harvest residue plants, assumed to be located in
Disposal of ash: Evidence from two Canadian an existing small urban setting, power plant sta?
plants is that once a biomass power plant starts compensation is estimated at $27 h1 to cover
up, a demand develops for ash, in that farmers salary plus bene>ts. For the remote whole forest
and foresters will remove ash from the plant case, a premium of 20% on all labor is applied.
at zero cost, and spread it on >elds [51]. How- Direct operating labor: A single boiler unit
ever, since this takes some time to develop, in requires eight operators per shift, and each
this study we have taken a more conservative additional unit requires an additional four
approach: ash is hauled to >elds at an assumed operators [49,51]. These levels are slightly
average haul distance of 50 km, and spread, all higher than comparable coal plants, and re-
at full cost to the power plant. For this scenario, <ect expected diPculties in the receipt and
spreading cost is 74% of total ash disposal cost. processing of biomass fuel.
Ash content varies for the three fuels, a?ected Administration costs: The biomass power
in part by the dirt content of the fuel. plant is assumed to be a stand-alone com-
Connection of the power plant to the existing pany, and an administration staPng level of
transmission grid: 26 is assumed for each case. In the whole
In the case of agricultural and forest harvest forest case, these sta? are sited at the remote
residues, the collection areas for biomass are location. If a larger >rm owned and operated
large, and there is some <exibility in the lo- the biomass power plant, savings in admin-
cation of the power plant, which is assumed istration costs would be possible. However,
in this study to be at or very near to an exist- these are not a signi>cant cost factor in the
ing community and to an existing transmis- overall price of power.
456 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

Maintenance costs: Maintenance is a major guaranteed return on investment. An alternate


source of uncertainty in evaluating biomass case is run at 12%.
plant operating cost. Existing power plants in
Alberta that pulverize and >re high ash coal
have maintenance costs in the range of $1.25 4. Key inputs
1:75 MWh1 . Various studies of biomass
units show values that are 710 times higher The key inputs for fuel properties, production
[43,49]. After some manipulation of actual and delivery of biomass, power plant characteristics
data from a small demonstration straw >red and costs, and general assumptions are presented in
power plant, we estimated maintenance costs Tables 14.
at about $13 MWh1 [52]. We cannot ex-
plain this wide range in terms of diPculty of
processing fuel or expected problems in the 5. Study results and discussion
boiler, and we attribute them in part to the
startup and demonstration nature of existing 5.1. Economic optimum size of power plant
plants. In this study we have assumed that
maintenance costs (parts plus labor) are 3% For the three sources of biomass, the economic
of the initial capital cost of the plant, which optimum size of power plant, the power cost and the
gives a maintenance cost in the range of geographical footprint from which biomass is
$4.93$6:20 MWh1 . Actual maintenance drawn are shown in Table 5.
costs in large-scale biomass facilities are a As expected, the economic optimum size of power
critical issue in overall economics of biomass plant based on biomass fuel increases with increas-
usage. ing biomass yield per unit area. Fig. 1 shows the
Plant reliability and startup pro3le: Biomass power cost as a function of plant capacity for the three
plants have operating outages that are often as- cases.
sociated with solids handling problems. In this These curves have two characteristics worth noting:
study, a plant operating availability of 0.85 is
assumed, which is less than levels of 0.90 0.95 The pro3le of power cost vs. capacity is ;at: In
routinely achieved in coal->red plants. Startup of biomass projects, two cost factors compete: fuel
solids based power generation is rarely smooth, transportation costs rise in approximate proportion
and this is accounted for by assuming a plant to the square root of capacity, while capital costs per
availability of 0.70 in year 1 and 0.80 in year unit capacity decrease. Because the variable com-
2. In year 3 and beyond the availability goes to ponent of fuel transportation cost becomes a signif-
0.85 [53]. The plants are assumed to be based icant cost factor as biomass yields drop, the result
loaded, which is a reasonable assumption in Al- is a very <at pro>le of cost vs. capacity. This result
berta where plants with a higher net marginal is consistent with previous studies of optimum size
cost (>red by natural gas) provide non-base load [15]. The <atness of cost vs. capacity for biomass
power. is di?erent than coal projects, where bigger is bet-
Reclamation: A site recovery and reclamation ter, and the size of a unit is often determined by
cost of 20% of original capital cost, escalated, either the largest available capacity or the largest
is assumed in this study, spent in the 30th year increment of power generation that the power mar-
of the project. Because the charge occurs only ket can accommodate. The result is that biomass
in the last year, it is an insigni>cant factor in the to power projects can be built over a wide range
cost of power. of capacities without a signi>cant cost penalty. For
Return: Power price is calculated to give a example, the economic optimum sized biomass
pre-tax return of 10%. The impact of rate of plant for whole forest is 900 MW (two maximum
return is assessed in a sensitivity case. This sized units), but the range of capacity for which the
value is consistent with a plant with a publicly power price is within 10% of the optimum value
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 457

Table 5
Economic optimum size of power plant for Alberta based biomass

Biomass source Biomass yield Optimum size Project area from which Power price
(dry tonnes per (MW) biomass is drawn ($ MWh1 )
gross hectare) (km)2
Agricultural residues 0.416 450 61,000 50.30
Whole forest biomass 84 900 19,000 47.16
Forest residues 0.247 137 764,000 63.00

Plant Size vs Power Price


Power Price (year 2000 US$ MWh-1)

90
80
70
60
50
40
Whole Forest
30
20 Forest Residues
10 Straw
0
0 500 1000 1500 2000 2500

Plant Size (MW)

Fig. 1. Power cost as a function of capacity for three biomass fuels

is 250 MW to more than 4000 MW. While the power price predicted to be within 10% of the
calculated optimum size for a forest biomass plant optimum value.
is 900 MW, in practice signi>cant road congestion
would occur at this scale, and the far more likely 5.2. The composition of power cost from biomass
plant size would be one 450 MW unit.
The assumption of maximum unit size drives the Table 6 shows the makeup of power cost per MWh
determination of the optimum size: The assump- for the three biomass cases. Note that costs are for the
tion that the largest single biomass unit that can >rst year of operation at full capacity (year 3), but are
be built is 450 MW puts a discontinuity in power de<ated back to the base year 2000.
cost at any multiple of that size, as is seen in
Fig. 1. This occurs because at 451 MW, two 5.3. Other important points
identical 225:5 MW units are built rather than
a single unit. For the straw and whole forest Several points are worth noting:
cases, the optimum size is found to be a multi-
ple of the maximum size of a single boiler. In Life cycle emissions of carbon: Table 7 shows
the case of straw, we looked at the optimum size the relative CO2 emissions per kWh for the three
of plant without the assumption of a maximum biomass cases used in this study and a new Alberta
unit size of 450 MW, and in this case the opti- coal >red power plant located at the mine. The
mum power plant size is 628 MW. However, as table uses the values of Spath et al. [66] for the
noted above, the <atness of the curve suggests construction of the power plant and the harvesting
that straw based power plants could be built in of biomass, and incorporates average haul distances
any scale from 145 to 900 MW with an output for biomass transportation. Transportation of coal
458 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

Table 6
Cost of power from biomass, year 2000 US$ MWh1 , at full capacity (year 3) and optimum size

Cost element Whole forest Forest harvest residue Agricultural residue

Capital recovery 16.97 20.72 16.32


Transportation 6.74 23.93 12.47
Harvesting 6.74 5.41 6.05
Maintenance 5.09 6.20 4.93
Operating 0.59 2.50 0.63
Administration 0.24 1.30 0.39
Field cost of biomass 2.45 2.30 2.29
Silviculture 1.39
Road construction 5.19
Nutrient replacement 6.64
Transmission 1.52
Ash disposal 0.25 0.64 0.57

Total 47.16 63.00 50.30

has a negligible carbon emission factor because in carbon credit required to make biomass power
western Canada the power plant is located adjacent competitive with incremental power from a new
to the mine. Emissions from transporting biomass coal >red plant, assuming a coal power cost of
are based on average haul distances for each spe- $30 MWh1 . A carbon credit of $18:30 tonne1 of
ci>c case. Note that even in the forest harvest CO2 , $21:70 tonne1 of CO2 , and $36:20 tonne1
residue case, transportation emissions are less than of CO2 would be required to equalize against an
5% of the emissions of a coal->red plant, per unit incremental coal plant for each of forest biomass,
of power. Emissions associated with mining coal straw and forest harvest residue. The Alberta power
are included, for both the energy required to move market was fully deregulated in 2000, and since that
the overburden and recover the coal, and the re- time monthly average power price has ranged from
lease of methane. Methane emissions from open pit less than $16 MWh1 to more than $165 MWh1 .
coalmines re<ect not only the methane contained in Fig. 2 shows the carbon credit that would be re-
the mined coal but also methane from the seam near quired to make the biomass cases economic in
the edge of the pit is released to the atmosphere. Alberta as a function of power price. These values
The approach of Hollingshead [67] was modi>ed to could be used to calculate a variable incentive for
re<ect the large size of a mine supporting a 450 or a publicly supported biomass power plant. Such an
900 MW coal >red power plant. Methane released incentive would be tied to actual power price rather
from the coal seam is estimated at three times than the cost of power from a displaced fossil fuel
the methane contained in the actual mined coal. plant i.e. by a new coal >red plant.
From Table 7 it is clear that this assumption does Whole forest biomass: Harvesting the whole for-
not signi>cantly a?ect the total estimated carbon est for power generation has the highest biomass
credit. yield per gross hectare and the lowest power cost.
Cost of biomass power: None of the biomass cases However, the cost is lowest only because nutrients
are directly competitive with coal-based power in are not replaced. Had the basis been nutrient re-
western Canada, which has a power cost (includ- placement, the cost of whole forest biomass would
ing return on capital) in the range of $30 MWh1 . be comparable to power from straw. The variable
Hence, in the absence of an emission credit biomass transportation cost is low, due to the high biomass
power will not be developed. From Table 7, the yield per hectare. Construction of roads is a major
di?erence in emissions is used to calculate the cost factor for power from biomass; the other two
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 459

Table 7
Life cycle emissions (g kWh1 ) from the power plants

Production Transportation Construction and Energy Total


decommissioning conversion emissions
Forest residues 28.0 [64] 35.5a 12.0 [64] 0 75.5
Whole forest 28.0 [64] 6.4a 12.0 [64] 0 46.4
Straw 28.0 [64] 8.9a 12.0 [64] 0 48.9
Coal 11.6c 0a 5.0 [66] 968.0b 984.6
a Based on truck transportation for a distance of 329, 52 and 123 km for forest residues, whole forest and straw respectively,

assuming the energy input of 1.3 MJ tonne1 km1 by truck and a release of 3 gC GJ1 km1 [65]. Most of the coal power
plants in western Canada are at a mine, so the transportation distance is very small. The emission during transportation would be
negligible as compared to the other components. Hence it has been neglected in this case.
b The emission factor is calculated based on characteristics of Alberta coal and the new 450 MW coal power plant.
c For Genesee, Alberta coal>eld [67]. It includes the contribution from methane emission and also the emission during the mining

of coal.

Cost of Power Cost vs. Carbon Credit


70
Whole
60 forest
Power Cost ($ MWh -1)

Straw
50
Forest
40 residues
30

20
10

0
0 10 20 30 40 50

Carbon Credit ($ tonne CO2-1 )

Fig. 2. Carbon credit required to make biomass power economic in western Canada as a function of average power price

cases utilize existing roads. This cost would dis- forest leads to a long rotation period, which in
appear for a second-generation power plant based turn gives a very low yield of residues per gross
on harvesting replanted forest. The remoteness of hectare. Areas that have shorter rotation periods
the assumed location for this plant is also a sig- would have more favorable economics for these
ni>cant penalty, giving a higher construction and residues.
operating cost and adding both transmission cost Agricultural residues: Power from straw is close
and line loss. If the whole forest plant were in a to whole forest utilization despite straws low yield
non-remote location, the cost of power would drop of biomass per gross hectare. Transportation costs
to $43:29 MWh1 . are high, but there is no cost for road infrastructure
Forest harvest residues: Forest harvest residues and the setting is not remote.
give the most expensive power of the three cases. Ash removal: Ash removal cost is based on the
The major cost penalty is the high cost of biomass conservative assumption of no credit for the nu-
transportation, which exceeds the cost of capital trient value of the ash; as noted above, there is
recovery. The slow growth of the northern Alberta evidence that once a biomass plant starts operation
460 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

that a demand for the ash emerges and that growers 6. Sensitivities
will haul it away at no charge to the plant. This is
evaluated in a sensitivity case. Some key sensitivities are shown in Table 8.

Table 8
Key sensitivities for power from biomass

Factor Power price Price impact Optimum size impact


($ MWh1 ) (%) (MW)

Capital cost of power plant is 10% lower


Agricultural residues 48.17 4.2 No change
Whole forest 44.96 4.7 No change
Forest residues 60.30 4.3 128
Pretax return on capital is 12% rather than 10%
Agricultural residues 53.71 +6.8 No change
Whole forest 50.70 +7.5 No change
Forest residues 67.27 +6.8 151
EPciency increased from 34% to 35% (LHV)
Agricultural residues 49.37 1.8 No change
Whole forest 46.46 1.5 No change
Forest residues 61.77 2.0 142
Largest unit size for boiler is unconstrained:
Agricultural residues (straw) case 50.03 0.5 628
Whole forest biomass location is not remote 43.29 8.2 450
Scale factor is 0.6 rather than 0.75
Agricultural residues 44.66 11.2 No change
Whole forest 41.14 12.8 No change
Forest residues 59.20 6.0 168
Scale factor is 0.9 rather than 0.75
Agricultural residues 58.42 +16.1 No change
Whole forest 55.52 +17.7 No change
Forest residues 66.47 +5.5 95
Biomass yield is 25% higher per gross hectare
Agricultural residues 49.29 2.0 No change
Forest residues 60.71 3.6 152
Whole forest biomass from
Good site (124 dry tonnes gross hectare1 ) 43.50 7.8 No change
Fair site (53 dry tonnes gross hectare1 ) 53.33 +13.1 No change
Biomass harvesting cost is 25% lower
Agricultural residues 48.78 3.0 No change
Whole forest 45.47 3.6 No change
Forest residues 61.64 2.2 No change
StaPng cost reduced by 25%
Agricultural residues 50.04 0.5 No change
Whole forest 46.95 0.4 No change
Forest residues 61.98 1.6 124
Ash disposal at zero cost
Agricultural residues 49.72 1.2 No change
Whole forest 46.90 0.6 No change
Forest residues 62.35 1.0 No change
Nutrient replacement, whole forest case 51.58 +9.4 No change
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 461

7. Other technologies and cost factors the power plant operator, security of fuel supply.
Failure to address reliability of fuel supply would
This study is based on production of electrical leave the power plant operator hostage to biomass
power from the direct combustion of biomass. Other price increases once the plant is built. This kind of
technologies warrant comment and future assessment: concern in the power industry is normally addressed
Use of low pressure steam for heating pur- by long term fuel supply contracts, which might
poses helps the economics of any thermal power work for forest biomass but would be unlikely to
plant project, i.e. biomass, fossil fuel or nuclear. work for agricultural residues coming from a wide
However, the potential for developing such a number of farmers. We believe that some social in-
co-generation application is higher for the agri- tervention will be necessary to address security of
cultural residues (where heat might be used in a fuel supply. For whole forest biomass, a key ques-
food processing application) and the forest harvest tion is the value of the wood as fuel vs. the value
residues (where the plant has such a large draw of the wood as >ber (lumber or pulp).
area that it might be economically located near This study assumes truck delivery of fuel. We note
a pulp or lumber operation). For a remote whole that no other power generation facility of signi>-
forest biomass, such a co-gen application is less cant size relies on highway truck delivery of fuel.
likely. Alternate transport mechanisms, includes rail and
In all cases of use of biomass fuel, water content pipeline, from hubs within the area from which
of the fuel reduces ePciency. This study does not biomass is drawn.
include an assessment of >eld drying of wood chips Direct combustion of biomass has a lower ePciency
or the use of very low quality heat, such as <ue gas, and lower heat rate than other technologies, notably
for drying of fuel. gasi>cation. Gasi>cation of wood can be achieved
Even though straw is passed through a combine as at signi>cantly lower temperatures (and hence cost)
part of the threshing operation, it is dispersed back than for coal [62].
onto the ground and then gathered up on a second
pass over the >eld. This practice simply re<ects past
agricultural needs rather than technological limits. 8. Conclusions
Dispersing straw on the >eld creates two problems:
there is extra cost for a second pass and the eP- Power from biomass in western Canada is not
ciency of straw collection is lower, since some of economic in its own right, but may become so if a
the straw, including virtually all of the cha?, is not system of trading GHG credits emerges. Whole forest
recovered on the second pass. biomass and straw can generate power for $47 and
For forest residues, a major cost is the forwarding $50 MWh1 at their optimum size. Forest biomass
(consolidation) and piling of residues prior to chip- likely requires a remote location with dedicated trans-
ping. If forest residue power projects were imple- mission, but has low transportation cost due to the
mented on a large scale, delimbing and topping of high biomass yield per gross hectare. Straw has a
trees could be integrated with chipping of the limbs lower biomass yield and hence a higher transportation
and tops in a single operation. cost, but this is partly o?set by a non-remote location
This study assumes that biomass fuel is sold at a and access to public road infrastructure. Forest har-
premium over cost of $4 dry tonne1 . An alterna- vest residues have a very low yield per gross hectare
tive for each of the fuels in this study is to require because of the long rotation and low cutting density
their availability at cost, as a condition of access in the boreal forest; transportation costs are of the
to Provincially owned timber in the case of forest same scale as capital recovery in this case, and the
harvest residue biomass, and as a condition of agri- cost of power is $63 MWh1 and optimum plant size
cultural support programs in the case of agricultural is the smallest, at 137 MW. Nutrient replacement was
residues. Such an approach would presumably re- factored into the straw case, but not into the forest
<ect a growing social concern for the need to miti- biomass cases since >rst cut operations in Alberta do
gate GHG. It would also address a critical issue for not practice nutrient replacement. However, repeated
462 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

forest harvesting ultimately requires nutrient replace- have not been reviewed or endorsed by any other
ment, and this is a signi>cant cost factor for the whole party.
forest case ($4:42 MWh1 ). Inclusion of nutrient
replacement makes forest biomass comparable to
power from straw. References
All biomass cases show a region of <at pro>le
[1] Jenkins BM. A comment on the optimal sizing of a biomass
of power cost vs. plant capacity, which occurs be- utilization facility under constant and variable cost scaling.
cause the reduction in capital cost per unit capacity Biomass and Bioenergy 1997;13(1/2):19.
with increasing capacity is o?set by increasing fuel [2] Nguyen MH, Prince RGH. A simple rule for bioenergy
transportation cost as the area from which biomass conversion plant size optimization: bioethanol from sugar
cane and sweet sorghum. Biomass and Bioenergy
is drawn increases. This means that smaller than
1996;10(5/6):3615.
optimum plants can be built with only a minor cost [3] Overend RP. The average haul distance and transportation
penalty. work factors for biomass delivered to a central plant. Biomass
Biomass yield per gross hectare is a major factor 1982;2:759.
in the cost of power from biomass, and forest harvest [4] Larson ED, Marrison CI. Economic scales for >rst-generation
biomass-gasi>er/gas turbine combined cycles fueled from
residue usage would be more economic in areas with
energy plantations. Journal of Engineering for Gas Turbines
shorter rotations. The assumption of maximum unit and Power 1997;119:28590.
size for a biomass boiler drives the optimum capacity [5] McIlveen-Wright DR, Williams BC, McMullan JT.
for the straw and whole forest cases, but the <atness of A re-appraisal of wood->red combustion. Bioresource
the cost vs. capacity curve means this is not a critical Technology 2001;76(3):18390.
[6] Dornburg V, Faaij APC. EPciency and economy of
factor. wood->red biomass energy systems in relation to scale
regarding heat and power generation using combustion
and gasi>cation technologies. Biomass and Bioenergy
Acknowledgements 2001;21(2):91108.
[7] Saskatchewan Agriculture and Food
Fibre Opportunity Agronomy Committee. Best management
The authors gratefully acknowledge the AMEC practices when harvesting surplus cereal straw, 2001. URL:
Corporation, the Poole Family and Bud Kushnir, http://www.agr.gov.sk.ca/docs/econ farm man/production/
whose >nancial support has made this research pos- creals/mgtpractices.asp
sible. Doug Topping, David Morrow, Fred Kardel [8] Carcajou Research Limited. Alberta annual report on small
and others at Epcor Corporation provided valuable areas, 1988.
[9] Hartman M. Estimating the value of crop residues. Agdex
assistance, including data on the historical cost of 519-25, Alberta Agriculture, Canada, 1999.
coal-based power and on transmission line capital [10] Canada Grains Council. Statistical Handbook, 1999.
and operating costs. David Williams, Chief Estimator [11] da Silva Pinto G. Spains largest straw->red power plant,
for Bantrel Corporation (an aPliate of Bechtel), and 1999. URL: http://www.nemesis.at/publication/gpi 99 2/
Wayne Silsbe, an estimator with AMEC Corporation, articles/<s.html
[12] Larsen JB. The worlds largest straw->red power plant,
provided valuable comments re capital cost estimat- 1999. URL: http://www.nemesis.at/publication/gpi 99 1/
ing. Wilsun Xu of the University of Alberta provided articles/27.html
insight into electrical transmission issues. Neil Shelly [13] Caddet Renewable Energy. The Masnedo CHP plantUsing
of the Alberta Forest Products Association provided indigenous CO2 neutral fuels. Technical Brochure No. 74,
contacts within the forestry industry who in turn pro- International Energy Agency (IEA), 1988.
[14] Caddet Renewable Energy. Straw->red CHP plant in
vided valuable insights into forestry practices and Rudkobingproviding environmentally. Technical Brochure
costs. Vic Lie?ers of the University of Alberta, Mike No. 95, IEA, 1998.
Apps of the Canadian Forestry Service, and Je? Chris- [15] Caddet Renewable Energy. The Sabro straw->red district
tiansen of Timberline Forest Inventory Consultants heating plant. Technical Brochure No. 88, 1998, IEA, 1988.
participated in valuable discussions regarding forest [16] Alberta Energy and Natural Resources. Alberta Phase 3 forest
inventory: yield tables for unmanaged stands, 1985.
biomass. Wayne Wasylciw of the Alberta Research [17] Puttock GD. Estimating cost for integrated harvesting and
Council provided input on straw usage. However, all related forest management activities. Biomass and Bioenergy
conclusions and opinions are solely the authors, and 1995;8(2):739.
A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464 463

[18] Hudson JB, Mitchell CP. Integrated harvesting system. [38] Desrochers L, Puttock D, Ryans M. The economics of
Biomass and Bioenergy 1992;2(1 6):12130. chipping logging residues at roadside: a study of three
[19] Hankin AC, Stokes B, Twaddle A. The transportation of systems. Biomass and Bioenergy 1993;5(6):40111.
fuel wood from forest to facility. Biomass and Bioenergy [39] Desrochers L, Puttock D, Ryans M. The economics
1995;10(23):14966. of chipping logging residues at roadside. EPciency and
[20] Hudson JB. Integrated harvesting system. Biomass and Alternative Energy Technology BranchCanada Centre for
Bioenergy 1995;9(15):14151. Mineral and Energy Technology, 1993, DSS Contract No.
[21] Perlack RD, Walsh ME, Wright LL, Ostlie LD. The economic 045SZ.23440-0-9639.
potential of whole-tree feedstock production. Bioresource [40] Desrochers L, Puttock D, Ryans M. Recovery of roadside
Technology 1996;55(3):2239. residues using drum chippers. Technical Report No. TR-111,
[22] Zundel P, Lebel L. Comparative analysis of harvesting and FERIC, 1995.
silviculture costs following integrated harvesting. Journal of [41] Hunt JA. Chipping roadside debris with the Bruks chipper in
Forest Engineering 1992;4(1):317. west-central Alberta. Technical Report No. SR-91, FERIC,
[23] Hall P, Gigler JK, Sims REH. Delivery systems of forest 1994.
arisings for energy production in New Zealand. Biomass and [42] Richardson R. Evaluation of Bruks o?-road chippers.
Bioenergy 2001;21(6):3919. Technical Report No. TR-71, FERIC, 1986.
[24] LeDoux CB, Huyler NK. Comparison of two cut-to-length [43] Bain RL, Overend RP, Craig KR. Biomass->red power
harvesting systems operating in eastern hardwoods. Journal generation. Conference Paper: Biomass usage for utility
of Forest Engineering 2001;12(1):539. and industrial power, Snowbird Resort and Conference
[25] McKendry P. Energy production from biomass (part 1): Center, Snowbird, UT, April 29 May 3, 1996, Engineering
overview of biomass. Bioresource Technology 2002;83(1): Foundation, NY.
3746. [44] OPce of Power Technologies, US Department of Energy.
Renewable energy technology Characterizations, Chapter:
[26] Zundel PE, Hovingh AJ, Wuest L, MacElveney D,
Direct->red biomass, 1997. URL: http://www.eren.doe.gov/
Needham TD. Silviculture systems for the production of
power/pdfs/direct >re bio.pdf
energy biomass in conventional operations in Atlantic
[45] Marrison CI, Larson ED. Cost versus scale for advanced
Canada, 1996. University of New Brunswick. URL:
plantation-based biomass energy systems in the US.
http://www.unb.ca/forestry/centers/biomass.htm
Proceedings of the Second Biomass Conference of the
[27] Silversides CR, Moodie RL. Transport of full trees over public
Americas: Energy, Environment, Agriculture, and Industry,
roads in eastern CanadaA state of the art report. Forest
Portland, Oregon, August 1995. p. 127290.
Engineering Research Institute of Canada (FERIC), 1985,
[46] Williams D. Personal communication. Chief estimator for
Special Report No. SR-35, Energy from forest (ENFOR),
Bantrel Corporation (an aPliate of Bechtel), Alberta, Canada,
Project P-312.
2002.
[28] Zundel P. The economics of integrated full-tree harvesting
[47] Silsbe W. Personal communication. An estimator for AMEC
and central processing in jack pine. Special Report No. SR-37,
Corporation, Alberta, Canada, 2002.
ENFOR Project P-322, FERIC, 1986.
[48] Gibson Steam Generating Plant. Supercritical pulverized coal
[29] Mellgren PG. Predicting the performance of harvesting
boilers. URL: http://www.fwc.com/power/products/
systems in di?erent operating conditions. Special Report No.
[49] Broek RVD, Faaij A, Wijk AV. Biomass combustion power
SR-67, FERIC, 1990.
generation technologies. Background report 4.1 for the EU
[30] Desrochers L. Personal communication. FERIC, 2002. Joule II+ Project: Energy from biomass: an assessment of
[31] Morbark Inc. Personal communication, 2002. two promising systems for energy production. Report 95029,
[32] Wiksten NA, Prins PG. Cost estimates of forest biomass Department of Science, Technology and Society, Utrecht
delivered at the energy conversion plant. Project Report P-19, University, 1995.
ENFOR, 1980. [50] Miles TR, Miles Jr. TR, Baxter LL, Bryers RW, Jenkins BM,
[33] Folkema, MP. Handbook for small to medium size fuelwood Oden LL. Boiler deposits from >ring biomass fuels. Biomass
chipping operations. Technical Report No. HB-07, FERIC, and Bioenergy 1996;10(23):12538.
1989. [51] Matvinchuk D. Personal Communication. Plant manager
[34] Bowater Newfoundland Ltd. Harvesting forest biomass as an Whitecourt Biomass Power Generating Station, Alberta,
alternative fuel. Project Report P-191, ENFOR, 1983. Canada, 2002.
[35] Favreau JFE. In-woods chipping: a comparative cost analysis. [52] Caddet Danish National Team. Straw->red biomass plants
Technical Report No. TR-105, FERIC, 1992. in Denmark. Caddet Renewable Energy Newsletter, March
[36] Spinelli R, Hartsough B. A survey of Italian chipping 1997.
operations. Biomass and Bioenergy 2001;21(6):43344. [53] Wiltsee G. Lessons learned from existing biomass power
[37] Asikainen A, Pulkkinen P. Comminution of Logging Residues plants. National Renewable Energy Laboratory Report No.
with Evolution 910R chipper, MOHA chipper truck, and NREL/SR-570-26946, 2000.
Morbark 1200 tub grinder. Journal of Forest Engineering [54] Prairie Agricultural Machinery Institute. Use of straw as a
1998;9(1):4753. farm heating fuel. PAMI Project No. RP0594, 1995.
464 A. Kumar et al. / Biomass and Bioenergy 24 (2003) 445 464

[55] Resource EPcient Agricultural Production Canada (REAP). [62] Williams RH, Larson ED. Biomass gasi>er gas turbine power
Woodchip Combustion in eastern Canada. URL: generating technology. Biomass and Bioenergy 1996;10(23):
http//www.reap.ca/Reports/Woodchips%20combustion%20in 14966.
%20eastern%20canada.htm [63] Hursh K. Rising nitrogen Fertilizer Prices. Aginfonet.coms
[56] Balers and baler attachment performance reports. URL: AgInfoBits News, January 15, 2001.
http://www.agric.gov.ab.ca/esb/afmrc/sum4a.html [64] Mann MK, Spath PL. Net CO2 emissions and energy balances
[57] Kimmins JP. Forest ecology. New York: Macmillan of biomass and coal->red power systems. Proceedings of
Publishing Company, 1987. the Fourth Biomass Conference of the Americas, Oakland,
[58] Environmental Manual database. Environmental manual California, 29 August2 September, 1999. p. 379 85.
for power development. Developed by German Technical [65] Borjesson PII. Emissions of CO2 from biomass production
Cooperation (GTZ) with scienti>c support from Institute for and transportation in agriculture and forestry. Energy
Applied Ecology (OkoInstitute), 1995. Conversion Management 1996;37(6 8):123540.
[59] Ashmead Economic Research Inc. Industrial uses of [66] Spath PL, Mann MK, Kerr DR. Report No. NREL/TP-570-
agricultural >bres. Prepared for Alberta Agriculture Food and 25119, Life cycle assessment of coal->red power production.
Rural Development, 1996. NREL 1999.
[60] Bakker-Dhaliwal R, Bernheim LG, Summers MD, Yan [67] Hollingshead B. Report No. CI 8936, Methane emissions
L, Jenkins BM, Lee H. Rice straw harvesting and handling from Canadian coal operations: a quantitative estimate. Coal
for o?->eld utilization. Prepared for the USDA, 1999. Mining Research Company, Devon, Alberta, 1990.
[61] Borjesson P. Economic valuation of the environmental impact
of logging residue recovery and nutrient compensation.
Biomass and Bioenergy 2000;19(3):13752.

View publication stats