Fuel Processing Technology 134 (2015) 251–258

Contents lists available at ScienceDirect

Fuel Processing Technology

journal homepage: [Link]/locate/fuproc

A modelling approach for the assessment of an air-dryer economic

feasibility for small-scale biomass steam boilers
L. De Fusco a,⁎, H. Jeanmart a, J. Blondeau b
a
Institute of Mechanics, Materials and Civil Engineering (iMMC), Universite catholique de Louvain, Place du levant 2, 1348 Louvain-la-Neuve, Belgium
b
Laborelec, Rodestraat 125, 1630 Linkebeek, Belgium

a r t i c l e i n f o a b s t r a c t

Article history: Fuel drying is an energetically and economically expensive pretreatment process, which may not be worth the
Received 2 October 2014 investment in the case of small-scale generation plants. This paper presents an investigation on the air dryer
Received in revised form 24 January 2015 feasibility to enhance the operation of biomass steam boiler. In the proposed approach, the external drying
Accepted 31 January 2015
technology using preheated air and the biomass steam production system is modelled in terms of energy and
Available online 13 February 2015
an economical analysis. A focus is given to the system size influence on the dryer economic suitability: the
Keywords:
smallest size of the biomass combustion system for which fuel drying is a suitable solution, from the economic
Biomass point of view, is computed. In the computations, the heat used for drying is assumed to be part of the cost for
Moisture operating the dryer and the thermal balance of the system is assumed to be previously verified. According to
Air-drying the model results, if the steam production plant operational time is above 8000 h/y, wood chips drying is feasible
Small-scale if the system size is larger than 1.78 tdaf/h of fuel processed.
Feasibility © 2015 Elsevier B.V. All rights reserved.

1. Introduction In most cases active drying is a necessary pre-treatment process

related to the technologies of thermal conversion of biomass: a high
1.1. Technical context moisture content decreases the efficiency of the energy conversion,
since the moisture must be first evaporated. To guarantee combustion
The biomass feedstock for energy production often contains a high quality, some industrial boiler technologies require the minimum low
moisture fraction: freshly cut biomass can include up to 65%w,wb. heating value (LHV) of biomass fuel to be above 15 MJ/kgwb. [3]. More-
(percentage on weight, on wet fuel basis) moisture when harvested, over, the auto-thermal and self-supporting combustion limit, for most
depending on the type of biomass and the environmental conditions. biomass fuels, is around 65%w,wb. of water content [5,6].
External (or surface) moisture is the moisture fraction above the If the fuel moisture fraction is below the auto-thermal self-
equilibrium moisture content and it generally resides outside the supporting combustion limit, drying is not necessary for combustion
biomass cell walls. Inherent moisture, on the other hand, is absorbed (in grate furnaces), but results in demonstrated benefits [1,3,4,7], such
within the cell walls. When the walls are completely saturated, the as:
biomass is said to have reached the fibre saturation point, or equilibrium
moisture [1]. The fibre saturation point increases with the relative • Increased and homogenised fuel LHV, with a decrease in the require-
humidity of external air and lower temperatures (Fig. 1) [2]. In wet ments for the combustion air pre-heating and process control.
and cold climates, the inherent fuel moisture might be as high as • Increased flame temperature, hence a potential increase in the steam
30%w,wb. [2]. As a consequence, open air drying is not effective. Despite production in existing facilities. According to Van Loo [3], if the
unforeseeable climate conditions (and logistical problems) leaving biomass fuel is dried from a moisture content of 50%w,wb. to 30%w,wb.,
wet biomass outdoors can have at times a positive effect concerning the boiler thermal efficiency can potentially be improved by 8–10%.
moisture [3]. Biomass materials are also hygroscopic; even dried and A lower air excess ratio may also be used if a more complete combus-
stored, they can still absorb moisture from the atmosphere, until the tion is reached thanks to the higher temperature.
equilibrium moisture content is reached [3,4]. • In existing combustion facilities, a decrease in the amount of flue gases
passing through the boiler (smaller emissions control equipment) and
lower product gas velocities (potential decreased erosion). In case of
⁎ Corresponding author. Tel.: +32 10472219. a new plant design, smaller heat exchange surfaces are needed
E-mail address: [Link]@[Link] (L. De Fusco). (decreased boiler dimensions).

[Link]
0378-3820/© 2015 Elsevier B.V. All rights reserved.
252 L. De Fusco et al. / Fuel Processing Technology 134 (2015) 251–258

oxygen and high temperature [12,14]. Air drying is the most flexible
solution and the process efficiency can be improved using multistage
drying [15]. Steam drying is sensitive to the material size and size
uniformity and, in general, despite the low heat specific consumption
(kJ per kg of water evaporated), is more expensive. In case of superheat-
ed steam dryers, the latent heat of vaporization and the water is easy to
recover and treat because the water vapour is not diluted with air [14].
Finally, the drying technology can be chosen (e.g. rotary dryers, belt
dyers, flash dryers), taking into account that each technology has
specific technical limits (e.g. the operating temperature) [14,16–19].
In conclusion, the choice of drying or not the fuel for power produc-
tion is not a trivial task. In case of larger biomass combustion units
(above 30 MWth) it has been experienced that it is worth paying the
investment for a feedstock dryer to improve the steam production
efficiency. A moisture content of 35%w,wb. is a trade-off between the
Fig. 1. Equilibrium moisture content of wood as a function of relative humidity and
enhanced combustion performance and the increased capital cost [20].
temperature (moisture at the finer saturation point), computed using the correlations
proposed in the Wood Handbook [2]. The trade-off between the dryer investment and the operating costs
on one side, and the enhanced boiler efficiency on the other side, in
case of small-scale distributed combustion systems is seldom explicitly
• In case of long term storage, reduced risk of biological contamination studied in the literature.
problems.
1.2. Literature review

These potential advantages are counterbalanced by numerous issues Brammer and Bridgwater [21] focused on the influence of feedstock
[4,5,8–10]: drying on systems coupling a small-scale biomass gasifier and an inter-
• Dryers have a relatively high investment cost, which largely impacts nal combustion engine. The minimum cost of electricity produced was
small-scale plants. Globally, though the boiler thermal efficiency determined incorporating a rotary dryer with a burner, drying from an
increases, the costs savings remain moderate [3]. Moreover, drying initial moisture content of 50%w,db. (percentage on weight, dry basis)
is an energy intensive operation. to a final moisture content of 10%w,db.. The highest overall energy effi-
• Explosions could occur and fires might arise if fuel ignition is reached ciency was obtained with drying to a final moisture content of 35%w,db..
during drying operation. Fire protection systems are necessary, Gebreegziabher et al.'s [9] approach was to maximize the annual
increasing the system investment costs. profit of the operation of the dryer for biomass combustion, without
• The dryer effluents have to be treated as exhaust gases or as wastewa- modelling the steam plant. The optimum solution indicated that the
ter discharges. dryer subsystem is profitable with the moisture level of the dried
• Burning very dry biomass (e.g. b 10%w,wb.) may increase the CO and wood at 17%w,wb.. Moreover, when the size of the wood chips becomes
the total particulate emissions [11]. Furthermore, the boiler opera- too large, the drying time is too long, thus significantly increasing the
tional temperatures, if increased, can approach the fusion tempera- dryer size and energy cost.
ture of some of the fuel ash constituents, increasing the slagging risk. Ho Ting Luk et al. [10] investigated how drying affects the overall
• If the boiler is designed for processing dry fuel and the dryer fails, the energy efficiency of a 12.5 MW biomass power plant that burns Empty
boiler becomes undersized for burning wet fuel. A backup low- Fruit Bunch (EFB), a feedstock with 60%w,wb. moisture, to support proper
moisture fuel may be needed. heat integration between the dryer and the power plant. In their study,
• Wet fuels can be utilized in grate combustion boilers, by increasing two types of dryers (which have different operating temperatures), a
the combustion air preheating (e.g. 300–350 °C) or by means of prop- Hot Air Dryer and a Superheated Steam Dryer, are proposed for the
er Flue Gas Recirculation (FGR). However, grates capable of handling drying process. With proper heat integration, the overall efficiency of
high combustion air temperatures, advanced air preheaters and excel- the production plant could be improved by about 5% when compared
lent refractories are necessary [4]. to process without drying. The economical aspects of drying were not
investigated.
In [13], H. Li et al. presented the results of a model of a belt conveyor
Additionally, the use of different drying media determines benefits drying system for pine wood chips at 60%w,wb. inlet moisture content,
and concerns (Table 1). Looking at the dryer energy efficiency and with flue gas and steam as drying agents, that could provide dried fuel
energy integration in the plant, flue gas drying and steam drying are to a 40 MW plant. While using flue gases as the heat source for drying,
to be preferred. However, with respect to air drying, both technologies the dryer capital cost computed was about €2.5 million; while using
might have lower capacity and control characteristics. superheated steam, the capital cost was about €3 million, because of
The use of low-enthalpy flue gas to dry the fuel is usually the less the higher quality steel in the equipment. The evaluation of the dryer
expensive solution, but fuel ignition is a risk in presence of sufficient profitability was performed defining a fuel-selling price, after the
dryer, without modelling the successive energy conversion. At a dried
Table 1 fuel selling price of 14 €/MWh, 3–4 years of operation was expected
Multi-criteria analysis for different drying media available [8,11,12]. Configurations where to give a return on the initial dryer investment.
different drying media are coupled are also commercially available. H. Holmberg and P. Ahtila in [22,23] have compared the exergy
Air Steam Flue gases efficiencies and the drying costs (capital and operational) of two types
of drying systems: single-stage drying with partial recycle of spent air,
Energy efficiency and heat integration − + +
Dryer effluentsa − + + and multi-stage drying. According to the results, the irreversibility
Fire and explosion risks − + − production depends to a considerable extent on the heat source and
Fuel size flexibility + − + the drying system. The single-stage drying is usually more economic
Control and capacity + − − when the amortisation time is short. However, the competitiveness
a
Dryer emissions depend also on the biomass type [13]. of multi-stage drying improves as the amortisation time becomes
 L. De Fusco et al. / Fuel Processing Technology 134 (2015) 251–258 253

Fig. 2. Single stage air drying system with partial recycling of exhaust (dm: drying medium, R: recirculation factor). Adapted from [15].

longer. In his PhD Thesis [15], H. Holmberg presents the heat specific in terms of energy and an economical analysis. The choice to model an
consumption and specific irreversibility computations for multi- air dryer is related to its flexibility, to the medium availability, and to
stage air drying with partial recycle of spent air and a Heat Recovery the possibility to use both an external or internal thermal source to
Unit. The optimally designed dryer consists of two drying stages, produce the medium at the desired thermodynamic condition. To assess
drying the fuel to the final moisture content of 30%w,db.. the dryer suitability, the additional steam produced in the boiler, thanks
to the use of dried fuel, is evaluated in economic terms and compared
1.3. Objective of the study with the estimated total dryer costs. In the computations, the heat
used for drying is assumed to be a cost for operating the dryer. The
To contribute to the analyses developed in the previous investiga- thermal integration of the dryer and the steam production system is
tions, this paper presents a study on the dryer economic feasibility not investigated. The energy balance should be verified, in a previous
with a focus on small-scale biomass combustion systems. The plant step, depending on the configuration, on a case-by-case basis. Given
size range explored is 1–6 tdaf/h (dry and ash free) woody fuel at the the parameters, the key finding of the modelling approach is the
inlet of the dryer and combustion boiler (b 30 MWth). A dryer using smallest biomass combustion system for which an air-dryer would be
preheated air and the biomass steam production system are modelled economically feasible.

2. Methods
Table 2
Model variables and values for the drying sub-model for the base case analysis.
2.1. Economic feasibility evaluation
Input parameters

Medium inlet T1a [∘C] 15 The Net Present Value (NPV) of the dryer investment is used to
Medium inlet p1a [kPa] 101.32 evaluate the benefits from drying and it has been computed as:
Flow r. humidity RH1 [−] 0.50
Flow r. humidity RH5 [−] 1.00 X
k
Rsteam−incr;y −C O&M;dryer
Medium p. losses Δp [kPa] 1.50 NPV ¼ −C INV;dryer ð1Þ
Medium velocity v [m/s] 0.65 t¼0
ð1 þ iÞt
Recirculation factor Ra [−] 0.30
Fuel dry mass flow mw,1a [kg/s] 1.0
Fuel moisture (in) Mw,1 [% w,wb.] 60
where Rsteam − incr,y [€/y], is the revenue due to the increased steam
Hot source qh,s = Φ/mdma [kJ/kgdry,air] 130 production; CO &M,dryer [€/y] are the dryer operation and management
Thot − source [∘C] 150 costs; and CINV,dryer [€] is the capital cost of the dryer. Capital is
Heat hx. thermal losses l [%] 2 amortised over k years of operation at a nominal interest rate of i% per
Dryer th. loss Φl [%] 10
year (assumed by hypothesis in the range of 5–10%). The revenue due
Power consumpt. Φm + e [kW] 2.0
Heat hx. factors Ui [W/m2/K] 10 to the increased steam production is computed as:
Fuel T. dryer step Tw,2 [∘C] ΔT = 5
Rsteam−incr;y ¼ C steam ΔGsteam Δhsteam 3600 ffl} τ ð2Þ
Output parameters |fflffl{zfflffl} |fflfflfflffl{zfflfflfflffl} |fflfflfflffl{zfflfflfflffl} |fflffl{zffl
s
operation
|fflfflfflfflffl{zfflfflfflfflffl}
Dryer area Ad [m2] c€
MJ
kg
s
kJ
kg
h h
y
HRU area AHRU [m2]
Pre-heater area Ahx [m2]
CINV,dryer [€] where Csteam is the steam economic value; ΔGsteam is the variation in the
 
Dryer heat consumption kJ=kgH2 O steam production due to the use of dried fuel; h its enthalpy content and
Fuel moisture (out) Mw,2 [% w,wb.] τoperation is the operational time of the plant (dryer and combustion
Drying medium mdm [m3/s] system). The heat source characteristics for the dryer operation and
a
Values are varied in the parametric analysis. the economic parameters are detailed in the next section.
254 L. De Fusco et al. / Fuel Processing Technology 134 (2015) 251–258

Fig. 3. Overview of the system studied, with its boundaries.

The model considers an average of the results of computing the First, the cost of an industrial dryer can be expressed as a function of
dryer investment cost CINV,dryer [€] with two industrial dryer system the dryer area, Ad [m2] [21]:
cost functions with the related hypotheses, adapted from [15,21,23]:
0:863
C INV;dryer;1 ½€  ¼ 7820ð2:79Ad þ 52:2Þ : ð4Þ

C INV;dryer;1 þ C INV;dryer;2
C INV;dryer ½€ ¼ : ð3Þ The second dryer investment cost CINV,dryer,2 is evaluated considering
2 also some additional costs, by using a multiplying factor G ≥ 1, named as
the Lang Factor [23]. This factor G is set at 1.60 on the basis of the
following adding cost values: electrical (+ 0.10), instrumentation
(+ 0.10), lagging (+ 0.05), civil work (+ 0.15), installation (+ 0.20)
[23].
Table 3
Variables for the combustion and steam production sub-model. X
C INV;dryer;2 ½€  ¼ G C equipments : ð5Þ
Input parameters

Biomass fuel Wood chips The equipment cost is computed as follows [23]:
LHV [MJ/kgdaf] f (Mw,2)
C [% w,daf] 49.60 C equipments ½€ ¼ C conveyor þ C hx þ C cover þ C drym−duct þ C fan ¼
0:7 0:5 0:5 0:7 ð6Þ
H [% w,daf] 6.20 ¼ 2700Ad þ 660Ahx þ 1200Ac þ 3770mdm þ 0:9Δpmdm
O [% w,daf] 44.20
S [% w,daf] 0.00
where Ad, Ahx, and Ac are respectively the conveyor dryer cross sectional
N [% w,daf] 0.00
area, the heat transfer area of the heat exchangers in the system (e.g.
Combustion heat recovery units, but not the dryer) and the dryer cover cross
Air inlet Tcomb − a [∘C] 25.0
sectional area. The height of the cover is assumed to be 6 m (industrial
Air inlet pcomb − a [kPa] 101.32
Fuel dry mass flow mw,2a [t/h] 1.0–10.0 dryers application). mdm is the drying medium mass flow [kg/s] and
Fuel moisture (in) Mw,2a [% w,wb] 10–60 Δp [Pa] is the drying medium pressure drop across the drying stage.
Comb. heat th. losses Φl,b [% LHVdaf] 5.0 The values of the area can be computed as a function of the flow
Air excess εb [−] 1.60 (density, velocity) or heat transfer (logarithmic mean temperature,
Recovery efficiency εhx [−] 0.90
Boiler approach point TAP [∘C] 20.0
heat exchange factor) variables from the dryer system model. A
Boiler pressure level (single) psteam [bar] 20.0 single-stage air drying system is considered in this study. It has lower
drying costs with respect to the more efficient multi-stage dryers [23],
Output parameters
consequently it is supposed to be a more suitable choice for small-
Limiting fuel moisture content Mlim [% w,daf]
Flue gas comp. [O2, N2, CO2, H2O, NO2] [% mol/mol] scale systems.
Steam mass flow msteam [kg/s]
Ideal combustion efficiencyc ηcomb − a [−] 2.2. Single stage drying system model
Ideal boiler efficiencyd ηboiler [−]
a
From the dryer model output. The single stage air dryer is modelled as a two-block system: a pre-
b
A single total air excess value. heater to produce the drying medium (heated from ambient tem-
c
Computed as the combustion products enthalpy content over the lower calorific value
of the biomass fuel entering the furnace.
perature using a hot primary source) [10,15] at the proper thermody-
d
Computed as the energy in the live steam, considering the reference for enthalpy at namic conditions and a drying chamber in which energy and mass
ambient temperature. exchanges between the fuel and the drying medium occur. The heat
 L. De Fusco et al. / Fuel Processing Technology 134 (2015) 251–258 255

Fig. 4. Operational map with the dryer specific heat consumption as a function of the air flow at the pre-heater inlet, of the pre-heater heat flux qh,s and the fuel moisture at the dryer outlet
0.45–0.05 [% w,wb.] (black isolines). Operations for drying 1 kg/s of dry biomass fuel.

for processing the drying air could come from, for example, the boiler or, 2.3. Simplified model for the steam boiler
when available, from other processes. This heat source is modelled as an
operational cost. In order to evaluate the advantage of burning dry fuel in the combus-
In Fig. 2, the dry air mass flow is indicated with mdm. It is heated in a tion system, a simple combustion and steam production sub-model has
Heat Recovery Unit (HRU, 1–2) and in the main heat exchanger (3–4) been developed. The two sub-models are coupled as represented in
with the heat flux Φ. Hot dry air goes through in the drying chamber Fig. 3.
(4–5), where losses Φl, mechanical work Φm, and electrical work Φe The model parameters for the boiler sub-model are summarised in
are considered. Wet air can be partially recirculated (R is the recircula- Table 3. During combustion, the biomass undergoes internal drying,
tion factor), then sent to the HRU (7–8) and finally to the exhaust. The devolatilisation and char combustion. The devolatilisation and the
importance of the saturated air recirculation lies in three reasons: secondary combustion of the oxidation products can be modelled, in
increased energy efficiency, possible optimization of the factor R to a simplified approach, with chemical equilibrium computations, as re-
maximize the moisture removal from the fuel, and the issues of control ported in the specific literature [24–28]. In this work, the gasification
(e.g. fire hazards). However, in case of simpler systems with no recircu- product composition is computed by minimization of the Gibbs free
lation, the input factor R in the model can be set to 0%. energy of the chemical system. Carbon in the residual char is oxidized
The behaviour of the systems is studied with the software Engineer- into carbon monoxide and carbon dioxide. The volatile gas produced
ing Equation Solver©. The main equations are mass, energy and exergy are then completely oxidized in the secondary combustion zone, an
balances applied to the systems, adapted from [22]. The thermodynamic air excess zone. Flue gas recirculation, preheated combustion air and
and psychometric properties are evaluated with the specific software air staging below the grate are not investigated. The steam production
functions. The key variables considered are reported in Table 2. is estimated by considering that all the heat produced in the combustion

Fig. 5. Operational map with the capital cost of the dryer as a function of the air flow at the pre-heater inlet, of the pre-heater heat flux qh,s and the fuel moisture at the dryer outlet 0.45–
0.05 [% w,wb.] (black isolines). The dryer estimated investment cost increases for higher drying medium flow rates and higher heat flows to produce the drying medium supplied.
256 L. De Fusco et al. / Fuel Processing Technology 134 (2015) 251–258

Table 4
Increment in steam production as function of the fuel drying, according
to the simplified combustion model.

Drying (in/out) Steam prod. ṁsteam

0.60–0.40% w,wb. +7.4%

0.60–0.30% w,wb. +8.0%
0.60–0.20% w,wb. +8.7%
0.60–0.15% w,wb. +8.9%
0.60–0.10% w,wb. +9.1%

For a given drying air flow rate, the moisture removal from the fuel is
higher for high dryer specific heat requirements (Fig. 4). The heat
specific consumptions are minimized for high moisture removal
(Fig. 4). For medium heat specific consumption and air flows, the
computed air temperature after the drying step (state 5, Fig. 2(a)) is
in the range 30–70 °C.
Fig. 6. Drying medium mass flow required to dry the fuel (left) and dryer heat specific From Fig. 4, it is possible to assess that, given the final moisture
consumptions (right). These trends, representing specific quantities, are valid for all the
content after drying, optimal couples of variables can be selected to
system sizes analysed.
maximize the dryer energy efficiency. According to the model results,
the operational area that corresponds to low specific consumption and
is exchanged in the convective pass of the boiler. Losses within the heat irreversibility is defined by the limits 70–190 kJ/kg provided at the
transfer are modelled by setting a recovery efficiency of the combustion heat exchanger and 10–30 m3/s of drying air flow.
gas energy content (see Table 3). The capital cost of the dryer increases in case of high drying air flows,
The model is able to determine the moisture content limit for self- because the cost is related to the cross sectional area of the dryer (Fig. 5).
sustaining combustion, defined as the fuel moisture content for which For a given drying air flow rate, the cost decreases in the case of higher
the primary combustion step (on the grate) is no more endothermic. specific heat flows provided to the heat exchanger to produce the
The computed limiting moisture value is about 60%w,wb. with the fuel drying air. From Fig. 5 it is possible to assess that, given the final
input data selected. õmoisture content, optimal couples of variables can be selected to
minimize the dryer cost.
3. Results and discussion From the dryer simplified modelling results, two additional prelim-
inary conclusions can be drawn:
3.1. Modelling results for the single stage drying system
1. The dryer energetic efficiency rise, when increasing the size of the
system and the mass of fuel dried.
Operational maps of the single stage drying system have been built
2. The influence of the recirculation factor on the system performances
to present:
can be positive or negative, depending on the operating parameters.
An optimum recirculation factor R opt that minimizes the fuel mois-
• The heat consumption [kJ=kgH2 O evaporated] (Fig. 4), which is an
ture content at the dryer outlet (or the heat consumption) exists
indication of the thermal efficiency of the system, and the specific
and can be computed with the model.
irreversibility rate [kJ=kg H2 O evaporated] (not shown here), which is
a measure of the exergy efficiency of the dryer [22]. According to the model results (Fig. 6), the drying medium mass
• The estimated CINV,dryer [€] (Fig. 5). flow rates required to dry the fuel increase when increasing the mois-
ture removal. Additionally, higher moisture removal allows a decrease
in the specific heat consumption. The medium requirements are related
These parameters are presented as a function of the drying medium to the recirculation of the saturated air (set to a fixed percentage, see
flow rate mdm [m3/s], the specific energy flux to heat the air medium in Table 2), which reduces the intake of ambient air flow rate. Further-
the heat exchanger qh,s [kJ/kgdry,air] and the moisture of the fuel at the more, the air drying system specific heat consumption is sensitive to
outlet of the dryer Mw,2 [% w,wb.].
Table 5
Main parameters for the economic modelling. The CO &M,dryer are expressed as % of the CINV,
dryer.

Parameter S1 S2 S3 S4 S5
h i
C steam Mc€J 1.5 1.5 1.0 Variable 1.3

CO &M,dryer[%] 15 10 10 8 Variable
hi
τ operation hy 8000 8000 6000 5000 Variable

i [%] 10 5 5 5 5

Table 6
Lower limit of the industrial combustion system size [tdaf/h] for which the drying system is
suitable (payback time of 5 years).

S1 S2 S3 S4 S5
a
mlim 1.8 1.0 (2.0 c€/MJ) 1.2 See Fig. 10
(1.3 c€/MJ) 4.9
(1.0 c€/MJ)a
a
Fig. 7. Dryers investment cost as function of the system size. Larger than 6.0 tdaf/h.
 L. De Fusco et al. / Fuel Processing Technology 134 (2015) 251–258 257

Fig. 9. Advisory tool result (scenario S1).

Fig. 8. Scenario S1. Dryers cumulated NPV as function of the moisture removal from the
fuel.

In the scenario S5, the minimum system size is expressed as a function

the losses associated to the additional heat exchangers in the system of the cost of the heat source for drying and of the operation time.
(air pre-heater and HRU). Considering the results for the scenario S1, the cumulated NPV of the
The dryer investment costs are computed varying the fuel mass flow investment is reported as a function of the moisture removal from the
rate in the range 1–6 tdaf/h. The cost CINV,dryer increases in case of higher fuel in Fig. 8. For the hypotheses considered, the net present value is
moisture removal and is higher for larger sizes (Fig. 7). higher in case of an increasing moisture removal: it is better to dry as
much as possible.
For the following results, by hypothesis,2 a target final moisture con-
3.2. Modelling results for the steam boiler system tent of 25% w,wb. at the combustion chamber inlet (dryer outlet) is set.
These results evidence that there is a lower limit plant size for which
Confirming the expected behaviour, the ideal combustion efficiency the increment in the steam production does not compensate (for exam-
decreases when the moisture content rises, due to a dilution effect. ple, in a limit payback time of 5 years) the investment and operational
According to the simplified model, burning dry fuel allows a moderate costs to sustain the fuel drying (Fig. 9). This limit system size is repre-
increase in the steam production (b 10% relative steam production in- sented graphically as the intersection of the cumulated NPV curves
crement, when drying from 60 % w,wb to 10 % w,wb), as is reported in and the x-axis and it corresponds to the estimation of the smallest
Table 4. With the simplifications considered, the feed rate has no direct plant size for which the fuel drying is a feasible economic solution. As
influence in the relative steam production increment. Experimental an example, with the stated hypotheses, fuel drying is feasible choosing
data on the increased steam boiler efficiency due to lower moisture air drying, if the boiler system size is larger than 1.78 tdaf/h (scenario S1).
contents are available, for example in [20], and a reasonable agreement The different scenarios are explored, see Table 6. In the scenario S2,
is found. decreased dryer operation and maintenance costs and a smaller interest
The heat flux used for drying is not accounted for within the energy rate are taken into account. Decreasing the utilization factor of the sys-
balance of the combustion plant. The energy requirements of the dryer tem remarkably changes dryer feasibility evaluation (scenario S3). The
are evaluated in economic terms as operation (O&M) costs. The results model results are quite sensitive to the steam price setting, as presented
of the next section allow the assessment of the feasibility of the dryer in the scenario S4.
coupled with the steam boiler, as a function of the system size. In Fig. 10, the minimum plant size to recover the dryer investment is
computed as a function of the system operating hours. Decreasing the
3.3. Economic coupling of the two systems availability and utilization factor of the system would increase the size
of the smallest system for which drying is feasible. According to the
The increased steam production is evaluated in economic terms and model results, in case of a small scale biomass combustion boiler with
compared with the investment and operational costs of the drying fuel drying, high production availability must be ensured to recover
technology. The evaluation is performed without considering the costs the dryer investment in 5 years of operations (S5, Fig. 10). As an exam-
of biomass purchasing (the feedstock cost at the system inlet, can ple, with the stated hypotheses, for a system processing 1.5 tdaf/h of
depend on its moisture content). wood chips, an air dryer (O&M costs set at 8%) would be economically
If focussing on the economic analysis, the variations of some param- feasible if the system availability is higher than about 8000 h/y. The
eters previously introduced (steam selling price, operation and mainte- influence of the drying heat source cost can also be studied (Fig. 10).
nance costs, dryer operational time and interest rate) allow presenting If considering the maintenance costs fixed at 4% of the investment
different scenarios (S), as shown in Table 5. The steam economic value costs [13], low operational costs can be set if the energy for producing
is based on its energy content and is assumed by hypothesis1 in the the drying air is available within the system (i.e. in CHP configurations)
range between 1.0 and 2.0 c€/MJ. The dryer O&M costs (including the or if producing the heat for drying the medium is thermally convenient,
heat source for drying) are chosen as a fraction of its investment cost, moreover if waste heat is available. High operational costs have to be set
decreasing from S1 to S4. High O&M costs correspond to high cost of if waste heat sources are not available or high exergy energy has to be
the heat for drying production, or high price of the heat source for dry- used,3 i.e. with a decrease of the boiler steam output. However, the
ing. Low O&M costs correspond to a high availability of thermal sources precise estimation of this cost will depend on the specific configuration
to produce the drying medium (i.e. CHP or low-exergy process heat). to be analysed.
The dryer operation time chosen for S1 corresponds to a high dryer
and boiler availability factor and is assumed to decrease for S3 and S4.
2
In case of enhanced drying (below 10%w,wb.), the NPV could decrease, because of the
1
It can be computed, in a first estimation, considering the fuel cost in [c€/MJ] and an av- dryer emission issues and secondary effects (as the fire risk).
3
erage combustion efficiency. In that case, other possible solutions for using the heat source should also be discussed.
258 L. De Fusco et al. / Fuel Processing Technology 134 (2015) 251–258

directly covered by the flue gases or through steam extraction,

depending on the case-specific energy balance. Finally, additional devel-
opment would be reached if modelling other configurations for drying,
such as flue gas and steam drying, and with a proper estimation of the
costs thanks to updated functions.

Acknowledgements

We gratefully acknowledge the reviewers for their helpful com-

ments. Professor V. Verda and Professor R. Napoli (Energy Department,
Politecnico di Torino) are acknowledged for the support to develop the
study.

References
Fig. 10. Results of the modelling for scenario S5. High operational costs would correspond
[1] P. Basu, Biomass Gasification and Pyrolysis Practical Design and Theory, Academic
to low availability of thermal sources for drying the fuel.
Press, Burlington, MA, 2010.
[2] F. products Laboratory, Wood Handbook: wood as an engineering material, General
Technical Report FPL-GTR-190, United States Department of Agriculture, 2010.
4. Conclusions [3] S. Van Loo, J. Koppejan, The Handbook of Biomass Combustion & Cofiring, Earthscan,
London, 2008.
It is widely known that a reduced biomass moisture positively [4] L. Rosendahl, Biomass Combustion Science, Technology and Engineering, Woodhead
Publishing, 2013.
impacts the combustion system performances, but the need for dry [5] B.M. Jenkins, L.L. Baxter, T.R. Miles Jr., T.R. Miles, Fuel Processing Technology 54 (1)
biomass feedstock requires large energy and capital costs on small-to- (1998) 17–46.
medium scale biomass plants. The main objective of the study is to eval- [6] A. Demirbas, Progress in Energy and Combustion Science 30 (2) (2004) 219–230.
[7] R. Van Der Broek, Biomass and Bioenergy (1996) 271–281.
uate in which terms an air-dryer investment is pertinent for small-scale [8] H. Song, F. Starfelt, L. Daianova, J. Yan, Applied Energy 90 (1) (2012) 32–37.
biomass steam boiler. A model to evaluate the component economic [9] T. Gebreegziabher, A.O. Oyedun, C.W. Hui, Energy 53 (2013) 67–73.
feasibility was built. Two energy systems are modelled: the dryer sub- [10] H.T. Luk, T.Y.G. Lam, A.O. Oyedun, T. Gebreegziabher, C.W. Hui, Energy 63 (2013)
205–215.
model (with air as drying medium) and a steam boiler simplified sub- [11] H. Hartmann, Low emission operation manual for chimney stove users (technology
model. Thanks to the first one, the drying investment costs are assessed and support centre of renewable raw materials), FutureBioTec Final Workshop
with correlations from the literature and as a function of the operating Technologies for Clean Biomass Combustion, Austria2012.
[12] L. Fagernäs, J. Brammer, C. Wiln, M. Lauer, F. Verhoeff, Biomass and Bioenergy 34 (9)
parameters. With the second sub-model, the effects of moisture on the (2010) 1267–1277.
combustion behaviour and boiler operations are verified. Both models [13] H. Li, Q. Chen, X. Zhang, K.N. Finney, V.N. Sharifi, J. Swithenbank, Applied Thermal
can be applied to different feedstock characteristics. Engineering 35 (2012) 71–80.
[14] A. Amos, Report on Biomass Drying Technology, National Renewable Energy Labora-
The modelling approach allows the assessment of the plant size for
tory, 1998.
which the drying system is not economically affordable. The results [15] H. Holmberg, Biofuel Drying as a Concept to Improve the Energy Efficiency of an In-
are presented in five examples (scenarios) and point out the influence dustrial CHP PlantPh.D. thesis Helsinki University of Technology, Department of Me-
of the system size and of the plant availability on the dryer selection chanical Engineering, Espoo, 2007.
[16] M. Aghbashlo, H. Mobli, S. Rafiee, A. Madadlou, Renewable and Sustainable Energy
feasibility. Drying resulted in a suitable technical solution, even for Reviews 22 (2013) 1–22.
small-scale systems, if considering a high plant availability. According [17] S. Vijaya Venkata Raman, S. Iniyan, R. Goic, Renewable and Sustainable Energy Re-
to the model results (scenario S1), if the steam production plant opera- views 16 (5) (2012) 2652–2670.
[18] M. Stahl, K. Granstrom, J. Berghel, R. Renstrom, Biomass and Bioenergy 27 (6)
tional time is above 8000 h/y, wood chips drying is feasible with air (2004) 621–628.
drying if the plant size is larger than 1.78 tdaf/h of fuel processed. [19] J. Berghel, R. Renstrom, Biomass and Bioenergy 23 (2002) 103–112.
In this approach, the thermal integration between the drying pro- [20] A.V. Bridgwater, A.J. Toft, J.G. Brammer, Renewable and Sustainable Energy Reviews
6 (3) (2002) 181–246.
cesses and the steam power plant is not explored, and the heat source [21] A.V. Bridgwater, J. Brammer, Biomass and Bioenergy 22 (2002) 271–281.
for drying is modelled as a variable O&M cost in the economic assess- [22] H. Holmberg, P. Ahtila, Applied Thermal Engineering 25 (17–18) (2005) 3115–3128.
ment, as a function of the plant size. A thermal balance for a specific [23] H. Holmberg, P. Ahtila, Biomass and Bioenergy 26 (6) (2004) 515–530.
[24] R. Bauer, M. Glles, T. Brunner, N. Dourdoumas, I. Obernberger, Biomass and
small-scale biomass steam plant, without economic analysis, was devel-
Bioenergy 34 (4) (2010) 417–427.
oped by Ho Ting Luk et al. [10]. Illustrations for the dryer integration are [25] N. Paces, M. Kozek, Modeling of a grate-firing biomass furnace for real-time applica-
presented in L. Fagernäs et al. [12]. tion, International Symposium on Models and Modeling Methodologies in Science
and Engineering (MMMse 2011), Vol. 6, 2011.
Future investigations could improve the modelling approach by
[26] C. Yin, L.A. Rosendahl, S. K. Kr, Progress in Energy and Combustion Science 34 (6)
further describing the heat source for producing the drying medium. (2008) 725–754.
This O&M cost variable could be more explicitly computed as a function [27] R. Strzalka, T.G. Erhart, U. Eicker, Applied Thermal Engineering 50 (2) (2013)
of the plant outputs (heat, steam, electricity costs), depending on the 1418–1426.
[28] T. Klason, Modelling of Biomass Combustion in FurnacesPh.D. thesis Lund Universi-
system thermal configuration. If the drying heat is coming from the boil- ty, Lund, 2006.
er, this would allow to evaluate if the heat demand of the dryer can be