Journal Identification = CHERD Article Identification = 595 Date: May 17, 2011 Time: 5:43 am
chemical engineering research and design 8 9 ( 2 0 1 1 ) 742–752
Contents lists available at ScienceDirect
Chemical Engineering Research and Design
journa l homepage: www.e lsev ier .com/ locate /cherd
Base en
impro p
Defin
Enrique Pa
a Departmen x 607
Montréal, Q
b Natural Re es, Q
a b s t r a c t
The development of a base-case process is a fundamental step in an energy efficiency study to obtain reliable results.
However, this step is often overlooked and there are no clear guidelines for the systematic development of the base-
case. A methodology has been proposed to properly define and evaluate the complete process for a subsequent
in
a
te
is
ti
d
c
p
K
m
1. In
The analys
the subjec
(2007) perf
three proce
consisted o
and of com
outputs su
district hea
several ele
Browne (19
steam con
pinch analy
power hous
� This arti
∗ Correspon
E-mail a
Received
0263-8762/$
doi:10.1016/
-depth energy analysis. It consists of two stages: definition and characterization of the process, and benchmarking
nalysis. In this paper, the first stage is presented. The base-case should encompass the process and the utilities sys-
ms, i.e., steam and water, as they are the driving forces of the chemical transformations. A four-pronged procedure
proposed to properly define and characterize a process and its utilities: data gathering, master diagram construc-
on, utilities systems analysis, and simulation. Themain objective is to build a computer simulationmodel to provide
etailed information on production, distribution, utilization and post-utilization treatment of steam and water. Pro-
ess inefficiencies are also identified, such as the low condensate recovery or the presence of non-isothermal mixing
oints. The procedure has been applied to an operating Kraft pulping mill in Eastern Canada.
© 2010 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
eywords: Energy efficiency; Base-case process definition; Utilities systems; Simulation; Kraft process; Energy opti-
ization; Water optimization
troduction
is of the operation of the Kraft process has been
t of a number of recent studies. Klugman et al.
ormed an energy audit in an operating mill with
ss lines of different construction period. The audit
f identifying all energy and electricity consumers
paring the results between the three lines. Energy
ch as effluents, power generated, steam venting,
ting are also quantified. A detailed description of
ments required for an energy study is shown by
99). These elements include the identification of
sumers, efficiency of the equipment, overview of
sis and process simulation, control strategies, and
e operation. Turner (1994) shows several elements
cle has been submitted in two parts. Part II presents the Benchmarking analysis.
ding author.
ddresses: enrique.mateos-espejel@polymtl.ca (E. Mateos-Espejel), jean.paris@polymtl.ca (J. Paris).
13 September 2009; Received in revised form17 June 2010;Accepted22September 2010
required to perform a water audit. These elements include the
identification of water consumers, effluents producers, effi-
ciency of the equipment, and reutilization strategies. In these
works, some of the aspects required in the development of a
base-case process and evaluation of the energy efficiency are
highlighted. However, they consider water and energy inde-
pendently and fail to cast light on the interactions between
the different sections of the process and between the steam
and water systems. In addition, they do not present a struc-
tured approach for the base-case development and for the
evaluation of the different elements to be considered in an
energy andwater study. Several noteworthy studieshave come
from Scandinavia. Axelsson and Berntsson (2005) performed
a pinch analysis in a state of the art-mill. Axelsson et al. (2006)
described two typical Scandinavian mills that vary by the level
– see front matter © 2010 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
j.cherd.2010.09.012
case process development for
vement, application to a Kraft
ition and characterization�
Mateos-Espejel a,∗, Luciana Savulescub, Jean
t of Chemical Engineering, École Polytechnique de Montréal, P.O. Bo
C, Canada H3C 3A7
sources Canada, Canmet ENERGY 1615 Lionel-Boulet Blvd., Varenn
ergy efficiency
ulping mill. Part I:
risa
9, Sta. Centre-ville,
C, Canada J3X 1S6
Journal Identification = CHERD Article Identification = 595 Date: May 17, 2011 Time: 5:43 am
chemical engineering research and design 8 9 ( 2 0 1 1 ) 742–752 743
of water ut
nities for h
to study the
and proces
2004). How
of these m
focuses on
analyzing o
between th
The dev
model of a
mization a
model, refe
the actual
ing conditi
detailed an
process de
should not
der its util
results pro
always give
foundation
taken.
Guidelin
and areas o
liest stage o
ensure suc
tion of ene
to as bench
analysis w
commitme
A system
model that
process ben
a Kraft pulp
ization of th
benchmark
of this wor
ing measur
the formula
completed
The analys
into accou
that impac
(Mateos-Es
Improvi
become an
energy pric
industry, it
face of eme
demand fo
the Canadi
A wide v
the energy
The extent
reutilizatio
of methodo
(El-Halwagi
and Smith,
larger the a
themore en
Steamand
although th
2005; Leew
Thus
and
mo
duct
ent
ds la
s of
rang
rep
ced
confi
cy. A
luate
alysi
d on
ahlq
cas
ft p
an e
rade
ats t
cons
Co
aft p
ich w
edia
d or
core
erfor
are
(whi
su
emi
erm
e sp
from
rnt i
nic c
and
reco
liquo
in a
sche
Kra
t/d (a
ill us
ion a
nt bl
. Th
weak
r in
ed th
to ra
drive
perfo
f cyli
d the
nsis
ent
ilization with the objective of identifying opportu-
eat integration. These models have also been used
implementation of new evaporation technologies
s integration measures (Wising, 2003; Bengtsson,
ever, the methodology used for the development
odels is not presented. In addition, the analysis
ly on the energy aspects of the process without
ptions for water reutilization and the interactions
e steam and water systems.
elopment of a representative, reliable and focused
n operating process is a prerequisite to the opti-
nd fine tuning of its energy performance. This
rred to as base case process, should represent
process in its current configuration and operat-
ons. It should be able to support a rigorous and
alysis leading to alternative, energy enhanced
signs that can be implemented in confidence. It
contain unnecessary details which could hin-
ization without improving the usefulness of the
duced. This critical and preliminary task is not
n the importance it deserves even though it is the
of all process analyses that may be later under-
es and targets that identify process inefficiencies
f most likely gains should be formulated at the ear-
f a retrofit project in order to channel efforts and
cess of the development, assessment and selec-
rgy enhancement options. This task is referred
marking. A thorough and careful benchmarking
ill later reduce the deployment of resources and
nt of expenditures.
atic stepwise method to construct a base case
meets those criteria and to execute effectively the
chmarking step has been developed and applied to
ing mill in operation. The definition and character-
e base case are presented in Part I of this paper, the
ing analysis in Part II. The anticipated sequence
k, i.e., the identification of potential energy sav-
es, their technical and economic evaluation and
tion of an implementation strategy has now been
and presented by (Mateos-Espejel et al., 2010b).
is has been based on a novel method which takes
nt the interactions and synergies of all systems
t upon the energy profile of the optimized process
pejel et al., 2010a,b).
ng energy efficiency in chemical processes has
important issue in times of volatile and increasing
es. In the case of theCanadianpulp andpaper (P&P)
is also part of a strategy to remain competitive in
rging pulp producing countries at a time when the
r paper commodities, the traditional mainstay of
an industry, is shrinking.
ariety of enhancing techniques is used to improve
efficiency of a feedstock transformation process.
of internal heat recovery and the degree of water
n are often the most important. A broad spectrum
logies has been developed to address the problem
and Manousiouthakis, 1989; Linnhoff, 1993; Wang
1994; Dhole, 1998). In a typical Kraft process, the
mount of water consumed and effluent produced,
ergy is required for heating, cooling, and pumping.
water systems are usually analyzed independently,
ey are strongly interconnected (Savulescu et al.,
ongtanawit and Kim, 2008; Mateos-Espejel et al.,
2008).
steam
The
on pro
treatm
depen
source
cover a
so as to
introdu
torial
efficien
to eva
for an
depen
2006; D
The
ing Kra
part of
high g
and tre
under
2.
The Kr
by wh
interm
finishe
2002).
The
step p
fibers
agent
sodium
and ch
and th
that th
rated
and bu
inorga
bonate
of the
green
on site
plified
Fig. 1.
The
700ad
The m
nificat
differe
ditions
of the
(DSC) o
is pass
steam
trains
ing is
a set o
ing, an
pulp co
equipm
, the focus of the base case model is the study of
water systems.
del developed must provide detailed information
ion, distribution, utilization and post-utilization
of those utilities. The reliability of the base case
rgely on the data used for its definition. Several
information must be consulted and the data must
e of operating conditions (e.g.winter and summer)
resent the main process variations. Simplifications
in the model must not modify the whole or sec-
gurations of the process which affect its energy
computer simulation must be designed as a tool
improvement scenarios and as a source of data
s (Lundström et al., 2007). Its level of details will
its main purpose (Turon et al., 2005; Blanco et al.,
uist, 2008)
e study presented below is based on an operat-
ulping mill located in Eastern Canada. The mill is
co-industrial cluster and, in addition to making a
Kraft pulp, it exports steam to a nearby sawmill
he effluent of an adjacent town; district heating is
ideration.
ntext
rocess is the prevalent manufacturing technology
ood chips are transformed into paper pulp, the
te material from which a very broad spectrum of
semi-finished paper products are made (Smook,
of the Kraft process is a chemical delignification
med in a digester where the individual cellulosic
separated to form the pulp. The delignification
te liquor) is a mixture of sodium hydroxide and
lfide. After delignification the fibers are washed,
cally bleached. Finally they are drained, pressed
ally dried. A key characteristic of the process is
ent delignification liquor, the black liquor, sepa-
the fibers in the washing step, is concentrated
n the recovery boiler to produce steam. The spent
hemicals form a smelt, composed of sodium car-
sodium sulfide, which is collected at the bottom
very boiler. The smelt is dissolved to form the
r which is recaustified with quick lime produced
lime kiln, to regenerate the white liquor. A sim-
matic of the complete Kraft process is given in
ft mill studied has an average production of
dt = air dried tons) of high grade bleached pulp.
es an 8 batch digester sequence for chemical delig-
nd a five stage bleaching sequence which uses
eaching agents (ClO2, H2O2, NaOH) at different con-
e ClO2 is manufactured on site. The concentration
black liquor (BL) initially at 15% dissolved solids
suspension is performed in two steps; first, the BL
rough a set of pre-evaporators driven by recycled
ise its DSC to 19% and then it is sent to two parallel
n by live steam to reach a final DSC of 75%. Dry-
rmed in two steps: first, the pulp passed through
nders where water is evaporated by indirect heat-
n hot air is used to attain the final specification of
tency. A steam turbine is used to entrain the drying
.
Journal Identification = CHERD Article Identification = 595 Date: May 17, 2011 Time: 5:43 am
744 chemical engineering research and design 8 9 ( 2 0 1 1 ) 742–752
e Kra
3. Me
The object
tion requir
be used fo
go beyond
steam audi
ity consum
process eva
to water or
tion. The m
steam toge
critical to a
ment, defin
of steam an
puter simu
order to ob
for a comp
The bas
steps:
• Data gat
• Master d
• Systems
• Simulati
The sys
tify ineffici
benchmark
3.1. Da
The data t
water beha
(i.e., steam
mation can
of the mill
instrument
contain de
streams (i.
data collec
tions signifi
pulp produ
ering the d
is perform
mation. Th
post-utiliza
tion 3.3. Th
den
ing c
l pro
diff
pti
ust b
Ma
ter d
l sig
ted f
orm
s sec
and
s dia
mpu
nne
uct t
te fa
res s
it op
Sys
ystem
rme
iliza
e v
ed at
vario
ct or
enta
Fig. 1 – Simplified diagram of th
thodology
ive of this methodology is to detail the informa-
ed for the development of a base-case that will
r a water and energy study. The aspects treated
what is commonly encompassed in an water and
t. Audits usually focus on the identification of util-
ers and producers without specifying the type of
luation that will be perfomed. Audits are applied
steam but rarely include both systems in conjunc-
ethodology presented deals with the water and
ther, and in addition it integrates other subjects
n energy efficiency study: data collection and treat-
ition of the process boundaries, characterization
d water networks and, the construction of a com-
lation. These issues must be reigorously treated in
tain a reliable simulation model that can be used
lete and detailed evaluation of the process.
e process simulation model is constructed in four
hering
iagram
analysis
on
tematic application of these steps will help iden-
encies in the steam and water systems prior to the
ing analysis which will be the object of Part II.
ta gathering
o be collected should represent the thermal and
to broa
operat
ationa
The
consum
tion m
3.2.
A mas
and al
extrac
to perf
proces
loops,
Thi
the co
interco
constr
evalua
measu
and un
3.3.
Both s
be pefo
and ut
Som
record
• The
• Dire
• Perc
viour of the process over a long period of time
and water consumption for one year). This infor-
be obtained from the data acquisition system
, from arechived data, and from the process and
ation diagrams (PIDs). PIDs are very useful as they
tails of the individual process units and process
e., temperatures, flowrates, concentration). The
ted should be particularly detailed for unit opera-
cantly affected by seasonal variations, changes in
ction or recurrent technical problems. After gath-
ata, a preliminary overall water and steam balance
ed to identify inconsistencies or a lack of infor-
e specifics about the production, utilization, and
tion of steam and water are discussed later in Sec-
e utilization of several sources of information helps
• Water te
• Water re
• Inefficien
out preh
instead o
• Sections
This inf
of details re
process.
3.4. Co
The proces
tive to stu
ft process.
the scope of the subsequent analysis as different
onditions can thus be evaluated and possible oper-
blems pinpointed.
erence between the data on steam production and
on, and between water intake and water consump-
e assessed to detect possible gross errors.
ster diagram
iagram where the steam and water utility systems
nificant process streams are clearly identified is
rom the PID’s. The AUTOCAD software can be used
this task. The diagram contains the details of all
tions, and major unit operations, the recirculation
connection between sections.
gram is an essential tool for the development of
ter simulation, as all process sections and their
ctions are identified. In addition, it will be used to
he flow diagrams of the utilities systems, and to
ctors affecting the feasibility of energy efficiency
uch as the distance between the process streams
erations.
tems analysis
s are defined in detail. Data reconciliation should
d when large differences exist between production
tion.
ery important characteristics of the utilities are
this stage:
us fuels used to produce steam
indirect heating
ge of condensate recovery
mperature levels
utilization strategies
cies such as the utilization ofmake-upwaterwith-
eating or utilization of pressure release valves
f cogeneration
with the highest consumption of water and steam.
ormation is used as a guide to determine the level
quired in the simulation in the different part of the
mputer simulation
s is simulated in a water-energy oriented perspec-
dy the interactions between the utility systems
Journal Identification = CHERD Article Identification = 595 Date: May 17, 2011 Time: 5:43 am
chemical engineering research and design 8 9 ( 2 0 1 1 ) 742–752 745
Table 1 –
and utiliz
Steam prod
Biomass
Fossil fue
Recovery
Recovery
Desuperh
Total
Steam utili
Total MP
Total LP
Total HP
Total
Production
and evalua
procedure
encompass
the definiti
steam syste
state and it
The obje
and water
flowsheet a
tems flows
process sec
energy effic
by a compa
A real p
ments of op
variations,
and water c
ues contai
imperfectio
lation shou
the real pro
4. Ca
4.1. Da
For this cas
tion: measu
and2005 fo
extracted fr
est consum
A low a
found for th
This differe
high cost o
good monit
applied to
between w
large differ
perhaps ju
however sin
process at a
perature. T
the mill en
PID’s and a
2 –
ation
wate
ted w
ened
l
utili
ted
ened
l
−util
gap
iliati
Ma
cific
hite
ater
ives
r of
The
am
is us
n. St
sh w
puts
satis
near
ed b
ng m
, wa
ma
s are
s ut
y po
uctio
Sys
Average mill data for the overall production
ation of steam.
Mass flow (t/h)
measured data
uction description
boiler (BI) 64.7
l boiler (FF) 29.5
boiler (RB1) 89.0
boiler (RB2) 37.5
eating water 55.7
276.4
zation description
85.2
170.2
32.4
287.8
−utilization −11.4 (4% diff.)
te potential energy enhancement measures. The
developed for the construction of the simulation
es: the modeling of the process unit operations,
on of the required level of details in the water and
ms, the convergence of the simulation to a steady
s validation.
ctive is to represent the unit operations as steam
consumers. The starting points for the simulation
re the process master diagram and the utility sys-
heets. The level of details used to describe specific
tions depends on their potential impact on the
iency of the process. The simulation is validated
rison between simulated and measured data.
rocess is never in a true steady state; local adjust-
erating conditions, equipment turn over, feed rate
etc. cause constant fluctuationswhich affect steam
onsumptions. Moreover, measured parameter val-
n noises or errors (random or gross) caused by
ns of sensors and recording equipment. A simu-
ld hopefully represent a long term average state of
cess.
se study
ta gathering
Table
utiliz
Input
Trea
Scre
Tota
Water
Trea
Scre
Tota
Input
fill the
reconc
4.2.
All spe
pulp, w
also w
Fig. 2 g
numbe
oped.
the ste
Water
ductio
the fre
The in
fuel to
from a
are dry
facturi
boilers
The
utilitie
proces
identif
constr
gram.
4.3.
e study there are two principal sources of informa-
red archived values for different years (2002–2003
r steam; 2006 forwater) and the PIDs. Thedatawere
om this sample for the two periods with the high-
ptions; winter for steam and summer for water.
nd very tolerable discrepancy of 4% (Table 1) was
e steamdata averaged over the 2005winter period.
nce is within the range of process variability. The
f steam may have been a reason to maintain a
oring of all steam users. The same procedure was
the water system (Table 2) but a difference of 34%
ater intake and consumption was observed. This
ence may be due to poor monitoring of water usage
stified by its low cost. This is not a good practice
ce very large quantities ofwater are used in a Kraft
temperature which is well above the intake tem-
he cost of water heating is a significant share of
ergy bill. An analysis of the water streams in the
comparison to the current practices were used to
The energy
proper man
in this secti
the distribu
within each
4.3.1. Ste
Initia