chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

journal homepage: www.elsevier.com/locate/cherd

Lead time estimation for modular production

plants

Stefan Sievers ∗ , Tim Seifert, Marcel Franzen, Gerhard Schembecker,

Christian Bramsiepe
TU Dortmund University, Department of Biochemical and Chemical Engineering, Laboratory of Plant and Process
Design, Emil-Figge-Str. 70, D 44227 Dortmund, Germany

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

Article history: Modular plant design is an approach for making chemical production more flexible and more
Received 23 February 2017 efficient. Different approaches for modular plant design have been developed, for example
Received in revised form 1 October in the CoPIRIDE or F3 factory project. They have in common, that lead time reductions for
2017 modular equipment are expected e.g. by utilizing design repetition or parallelization of pre-
Accepted 5 October 2017 assembly of modules. To support the decision for or against a modular concept, besides
Available online 13 October 2017 cost effects possible lead time changes compared to conventional concepts should be antic-
ipated in early economic evaluations already. In this article, a lead time estimation method
Keywords: will be presented that correlates project costs and project durations and can be applied to
Modular plant design modular and non-modular plants enabling comparative studies. An example from a previ-
Modularity ous paper was used to investigate the impact of modularization on lead time. It includes
Time-to-market modular production lines and a non-modular backbone facility that provides energy and
Lead time utility supply. A range of investment sizes (FCI of 3–95 mio. D ) was investigated and com-
Fixed capital investment pared with a conventional reference plant. Total lead time reduction was in the range from
2.6 to 5.5 month depending on investment size. For a more significant impact on the lead
time the modularization approach needs to be modified by also applying modular design to
the backbone facility. In this case depending on investment size total lead time reduction
would be between 3.9 and 18.7 months representing a very significant reduction of 23%–60%
compared to the lead time of the conventionally designed reference plant. This is consid-
ered as the maximum expectable lead time reduction that can be achieved through modular
plant design. This reduction would represent a major potential for speeding up construction
of chemical plants.
© 2017 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction kets and an increasing demand for a flexible production (Buchholz,

Future development of global chemical market is characterized by
diversification and fragmentation (Lier et al., 2015). Technological
improvements and new fields of application make customers ask for
more tailor-made products, which leads to an increasing number of
products, decreased production volumes, delocalized product demand
and shorter product life cycles. This in turn leads to more volatile mar-
2010). Additionally, there is increasing pressure on product prices and
a trend of increasing raw material prices in the long term. To keep
up with this future development, future chemical production will also
need to be more efficient.
Fulfillment of both, increased flexibility and efficiency is hardly
possible applying existing production concepts. In a simplified view
the existing production concepts can be described by the benchmarks

∗
Corresponding author.
E-mail addresses: Stefan.Sievers@tu-dortmund.de (S. Sievers), Tim.Seifert@tu-dortmund.de (T. Seifert),
Marcel.Franzen@tu-dortmund.de (M. Franzen), Gerhard.Schembecker@bci.tu-dortmund.de (G. Schembecker),
Christian.Bramsiepe@tu-dortmund.de (C. Bramsiepe).
https://doi.org/10.1016/j.cherd.2017.10.003
0263-8762/© 2017 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Fig. 1 – Scheme of the modular design of the F3 factory
concept.
of the large-scale commodities production plant and the small-scale
specialties production plant (Bramsiepe, 2014).
Traditionally, the main economic driver of the continuously oper-
ated large-scale commodities production plant is the economy of scale
together with high process efficiency. However, this concept usually
offers a low degree of product and capacity flexibility (Buchholz, 2010).
The small scale specialties production concept, e.g. a multi-purpose
batch plant, offers a higher degree of flexibility regarding capacity and
product portfolio. Here, the ability to quickly react to seasonal or new
market requirements including innovations are main economic drivers
(Rauch, 1998). The disadvantage of this concept is the comparably low
efficiency due to a lack of material and heat integration (Rauch, 1998).
A solution proposed to combine the advantages of both production
concepts are production plants that use a modular design (Buchholz,
2010). Several approaches to describe modularization are available.
While Jameson (2007) describes a module as just being a mobile unit, a
more detailed description has been developed by Burdorf et al. (2004),
Kampczyk et al. (2004) and Schmidt-Traub et al. (1999). In their defini-
tion a module corresponds to a main equipment item including its local
pipe installation. A module description that also includes standardiza-
tion of modules has been developed in several research activities in
the recent years. European research projects to be considered are for
example CoPIRIDE (Löb, 2013) or F3 factory (Buchholz, 2010; Buchholz,
2013), further relevant projects have been listed by Lier et al. (2015).
The concept used as a reference for modular design in this paper
was used in the F3 factory concept. F3 stands for flexible, fast and
future production plants and describes a modular and continuous
operating mode. The overall F3 factory plant is made up from two mod-
ular structures, PEAs (Process Equipment Assembly) and PECs (Process
Equipment Container), both equipped with standardized interfaces.
The structure is schematically illustrated in Fig. 1.
In this concept, modular equipment fulfilling a full unit operation
(the PEAs) can be placed into standard containers which represent the
modular superstructure (the PECs) and are connected to a shared back-
bone facility that provides housing and basic supply with utilities and
energy. Aiming for high process efficiency highly automated, contin-
uously operated processes using process intensified technologies and
material and heat integration can be implemented (Seifert et al., 2014;
Bramsiepe et al., 2012).
However, process intensified technologies are often limited in
throughput. Especially when parallelization or numbering up is
required to prepare technically relevant production capacities, loss of
economy of scale can compensate improved conversion cost.
An economic evaluation is necessary to determine whether the
modular approach provides economic advantages over conventional
production concepts.
For such economic evaluation it is not sufficient to incorporate
investment costs and conversion costs only. There are further eco-
nomic aspects like supply chain costs, personnel costs, site-specific
costs, and others that are affected by implementation of a modular
plant design and need to be considered (Sievers et al., 2017).
One of the most important effects on the economy is the increase
of plant flexibility. The structural flexibility obtained by the modular
design poses the chance to build processes quickly and in multiple
small scale units (Buchholz, 2010). From lab to production scale a small
factor applies, reducing scale-up issues. Potentially, existing modules
can already be used for product development and production of sample
amounts while production capacity can later be increase by number-
ing up. Such integration of product and process development, leading
to a reduction of development periods was shown e.g. by Brodhagen
97
et al. (2012). Another feature that can be exploited using modular-
ization is standardization. Standardized modules and standardized
simulation models can be used for faster process development by
selecting modules from a module database rather than designing them
individually. For example, in two studies (fine chemicals and pharma
product) Grundemann et al. (2012) showed that the time-to-market
can be reduced by implementing pre-designed, continuously operated
micro-structure units.
Moreover, pre-construction of modules in an off-site workshop
under ideal spatial conditions and the presence of all necessary tools
and expertise may lead to a reduction of field work but requires a high
degree of logistics planning, additional accommodation for workers
at the site and can potentially be dependent on weather conditions.
Nevertheless, construction expenses and time may be reduced.
The reduction of engineering and construction effort helps to
decrease lead times and thus the time to market, which can substan-
tially improve the economy by generating earlier sales. As could be
shown by Bramsiepe et al. (2012) and Seifert et al. (2012), a lead time
reduction from three to one year by means of modular setup imple-
mentation can result in a net present value improvement of more than
25 percent, which offers the opportunity to compensate loss of econ-
omy of scale. The generation of earlier sales also results in a shorter
pay-back period, reducing the investment risk. This in turn allows to go
for projects that otherwise were too risky, creating an increased degree
of business flexibility.
It becomes clear that lead time is a central element for economic
evaluation and comparison of modular and conventional production
concepts. There are different project phases, which constitute the lead
time. An example of a typical differentiation of lead time phases is
given, for example, by Bramsiepe et al. (2011).
The first economic evaluation that already includes a lead time
estimate is conducted during a feasibility study in the process con-
cept phase (Navarette, 2009). This evaluation aims at choosing one of
the process alternatives developed, e.g. a modular vs. a conventional
approach. Bramsiepe (Bramsiepe, 2014) found that for conventional
plants in this phase only a rough estimate is conducted for lead time
estimation because “especially during early process design only few informa-
tion about the process under investigation are available so that the decision pro
or contra a process alternative has to be made based on practical knowledge
instead of detailed calculations”. For example, in this phase Mosberger
(2012) propose to estimate two to two-and-a-half years from process
design freeze to start-up plus half a year to one year until minimum
capacity is reached that corresponds the break-even point.
Even rough lead time estimates need to be established on a com-
mon and comparable basis to get meaningful results. Thus, comparing
modular and conventional design requires that the same methodologi-
cal approach is applied and the same database is used for the lead time
estimate of both approaches. Of course above mentioned expectable
lead time reductions for modular plants need to be considered in such
comparison. This means existing rough lead time estimates for conven-
tional plants need to be adjusted and for plants with modular design a
new estimate is required.
2. Approach
For lead time estimation, first, the phases of an engineering
project have to be defined. After that a calculation method
has to be developed that allows determining the duration of
each of the phases.
For further considerations, it is assumed that there is a rela-
tionship between the duration of a project and the investment
sum as for any given project, accruing costs and the project
duration can be expressed as a relation, i.e. different durations
will yield different costs (Khosrowshahi, 2002).
As a starting point for the lead time calculation thus, the
fixed capital investment (FCI) must be calculated first. The
starting parameters used for calculating phase durations are
the expenses that accumulate by the engineering activities
98 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Fig. 2 – Illustration of the most important phases of an investment project and assigned project costs. The period
considered in the presented method is marked by grey background. Period durations are not true to scale.
The lower part of the picture was adapted from Bramsiepe et al. (2011).
and the construction work to be conducted in an investment
project. These parameters can be derived e.g. from existing
FCI calculations as those provide descriptive parameters for
investment projects. Additionally, an estimation of the FCI
including costs of construction work and engineering activ-
ities contributing to it has recently become available for both
conventional and modular plants using the same methodol-
ogy. This opens the opportunity to identify changes resulting
from implementing modular technology and their impact on
material, engineering and construction costs on FCI (Sievers
et al., 2017), which is the basis for correlating project expenses
and required project execution time.
Input variables of the proposed method are construction
and engineering costs. Using correlations for project staffing
and salary information the approach presented provides the
opportunity to convert expenses for construction work and
engineering activities into time intervals and thus to calculate
an estimated lead time.
2.1. Application of lead time estimation
In Fig. 2 typical phases of an investment project are illus-
trated for a conventional chemical production plant. In the
conceptual design phase, the best alternative among possible
production concepts developed is identified gradually, which
may include modular and non-modular production concepts.
In chapter 1 we have ascertained that in this phase a first eco-
nomic evaluation is necessary to compare these production
concepts and as the lead time is a decisive factor for the eco-
nomic result, an estimation of the lead time is also required
in the conceptual design phase.
In later phases of the project parallel work on several alter-
natives should be avoided. This is important, since the costs
increase with the progress of project duration and here paral-
lel work on alternative concepts that may be discarded later
is particularly costly (Baerns et al., 2006). The project costs
as well as the number of process alternatives applied to the
project phases are shown in Fig. 2. Starting with the basic
engineering phase the costs substantially increase. To include
these costs and the corresponding phase duration the basic
engineering costs should be included in the FCI estimation
and the lead time estimation as well; the estimated period
should include the phases from basic engineering to commis-
sioning and start-up and thus all phases after the estimate (see
shaded area in Fig. 2). The estimate itself should take place in
the concept phase.
2.2. Definition of phases in the lead time period
Fig. 2 shows that some phases of the lead time are conducted
in parallel. Actually, in literature the phases are separated and
described differently due to their complexity and mutual links.
Thus, a clear separation of such phases is already difficult
today and “will become even more difficult if lead times are reduced”
(Bramsiepe, 2014).
However, from a management point of view a stage-
gate process is necessary to approve breakpoints sequentially
(Peters et al., 2003). Referring to this in our approach a clear
specification and sequential separation of lead time phases is
developed. In the following, the lead time is broken down into
a set of sub phases based on various literature sources.
According to Helmus (2008) a clear separation into project
planning and project execution is possible. The project
planning phase can further be separated into product devel-
opment, process development and implementation phase
(Mosberger, 2012; Festel et al., 2001; Wagner, 2003; Sattler and
Kasper, 2000) (compare Fig. 3).
As described above the product development phase is not
included in the lead time. Process development in contrast
takes place during the conceptual design phase (compare
Fig. 2) and aims at selecting an optimum process alternative
including lead time estimation. Using that optimum alterna-
tive, a feasibility study can be conducted. Given a feasible
process alternative, the project execution phase is initiated
(Mosberger, 2012; Navarette, 2009). As explained above the
lead time estimation takes place during process development
phase and includes all phases after the estimate. Conse-
quently, the beginning of the lead time period to be estimated
is not identical with the end of the process development phase
or the start of the implementation phase respectively. For
 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Fig. 3 – Phase definition of an investment project as defined by Helmus (2008) (top row), Mosberger (2012), Festel et al.
(2001), Wagner (2003), Sattler and Kasper (2000) (middle row) and Navarette (2009) (last row). The lead time phases chosen
for the proposed approach correspond to the structure according (Navarette, 2009). Period durations are not true to scale.
clear phase separation a definition used by Navarette (2009)
is applied who defines a pre-project phase before the imple-
mentation phase. The pre-project phase represents a clear
starting point and is the first phase contributing to the lead
time. Following Navarette’s (2009) nomenclature the imple-
mentation phase as defined by (Mosberger, 2012; Festel et al.,
2001; Wagner, 2003; Sattler and Kasper, 2000) will be called
project start phase in the following (compare Fig. 3).
As a result of the process development phase at the begin-
ning of the pre-project phase basic knowledge about the
process is already available; e.g. reaction kinetics and a block
diagram of the process is available, from laboratory experi-
ments separation parameters including expectable yields are
known and market forecasts and an analysis of the patent
situation are available. Before the pre-project starts, a first
process flow diagram including equipment list and mass
and energy balance is available, too. Investment capital for
proceeding with the next project phase is approved after suc-
cessful check of economy i.e. the conceptual design is fixed for
the follow-up phases, beginning with the pre-project phase.
The pre-project phase is the first phase contributing to
the lead time and represents the transition from process
development to implementation phase. In this phase, basic
engineering is initiated and the production concept and pro-
cess design are refined. Results of this phase are (Navarette,
2009):
• Fixed plant location and size
• Fixed product specifications
• Fixed material and energy balance
• Definition of recycle and purge streams
• Definition of equipment size
• Raw material and utilities selection
• List of motors and actuators
• P&ID and preliminary floor plan
• Concept for instrumentation and automation
• HAZOP and environmental pollution study
• Packaging and logistic concept
Basic engineering work is ongoing at the end of the pre-
project phase but the results until this time are the starting
point for first detail engineering activities and mechanical
design (Navarette, 2009).
After completion of the pre-project, the “project start”
(Navarette, 2009) or “implementation phase” (Mosberger, 2012;
Festel et al., 2001; Wagner, 2003; Sattler and Kasper, 2000) is ini-
tiated. In this phase increased project planning, formation of a
99
project team and selection of contractors are conducted. Fur-
thermore, basic engineering activities are intensified, building
the basis for more detailed cost estimation (±10 % to ±25 %)
and the final approval for expenditure estimate (AFE). Prepa-
ration of AFE and a waiting time for final AFE approval for
construction start are also assigned to this phase (Mosberger,
2012). For the calculation procedure presented, we assume a
positive AFE decision, that investment capital is released and
project execution is initiated.
Project phases until this point are assigned to the project
planning period. The second superordinate phase is the
project execution period (Helmus, 2008). This period can fur-
ther be separated into the phases basic engineering, detail
engineering, procurement, assembly/erection and commis-
sioning (Madauss, 2000; Mosberger, 2012). In fact, first basic
engineering activities do start in the pre-project phase before
the basic engineering phase due to planning tasks that are
necessary as basis for later activities. The same holds true
for detail engineering and also for construction activities. For
example, while first site preparation and excavation work is
executed, detail planning of secondary equipment, piping, fit-
tings, etc. still needs to be elaborated (Sattler and Kasper,
2000). The period between AFE approval and first construc-
tion work incorporates preparatory work that mainly includes
detail engineering tasks (Navarette, 2009). There is a mis-
match between a strict differentiation of the basic and detail
engineering phase and the actual starting times of different
engineering activities to be considered. As a consequence, we
concluded that it is more practical to follow Navarette’s (2009)
approach who just divided the project execution phase into
an “engineering lead time” and the “construction” phase.
The engineering lead time follows the “project start” resp.
implementation phase (compare Fig. 3). Before construction
can start, preparatory work necessary for first construction
tasks needs to be completed. In parallel to this preparatory
work, the main part of detail engineering is finalized. During
the engineering lead time, the following actions are performed
to prepare start of construction:
• Finalize long delivery equipment specification and select
vendor
• Complete basic engineering
• Complete detail engineering to approx. 90%
• Contracting for site preparation, foundations and steel
structure and supports
Of course, engineering work is not finished after engi-
neering lead time. First construction work is performed
100 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
immediately after engineering lead time and is carried out
in parallel to final detail engineering activities (Navarette,
2009). Further engineering-related activities during construc-
tion phase are procurement, site supervision and enhanced
project management (Mosberger, 2012).
Whether commissioning resp. start-up is considered as
part of project execution or not is a question of definition.
While Bernecker (2001) or Sattler and Kasper(2000) summarize
technical acceptance testing and start-up as commissioning,
Navarette (2009) distinguishes between these activities and
considers the start-up as separate phase while all acceptance
tests are considered as part of construction finalization. By
choosing Navarette’s definition of the construction phase we
decided to consider start-up as separate phase, avoiding to
take acceptance tests into account twice. Thus, both engi-
neering lead time and construction together are considered
as project execution phase. The project phases defined for
further use are shown in Fig. 3.
3. Calculation procedure
According to our approach, the lead time duration (tLT ) is the
sum of the durations of the pre-project phase (tPP ), project start
phase (tPS ), project execution phase (tPE ) and start-up phase
(tSU ).
tLT = tPP + tPS + tPE + tSU (1)
As described in chapter 2 expenses for construction work
and engineering activities will be converted into time inter-
vals to determine the values of the summands. During project
execution phase calculation interim results are generated that
will be used for calculating other phase durations. Therefore,
the description of the phase durations’ calculation starts with
the project execution phase.
The project execution phase (tPE ) is the sum of the con-
struction phase duration tc and the engineering lead time teLT
(compare Fig. 3). First, the construction phase duration is cal-
culated.
The duration tc is calculated depending on the construc-
tion work hours hc . For this a result given by Navarette (2009)
is used who found a correlation between the actual construc-
tion working hours on the field and the actual field duration
i.e. duration of construction phase (see Fig. 4). This correlation
is not linear because first, a basic amount of working hours
is necessary nearly independently on the size of a project,
e.g. contractor’s job preparation scheduling, management of
workers and accommodation or equipment transport, and
second the bigger the project is, the more work is performed
in parallel.
For calculation of the duration of construction phase tc the
correlation by Navarette can be described by an approximate
function determined by least square method:

3 × 10−11 × hc + 5.53 × 10−5 × hc + 2.8283 ∀hc ≤ 105 ; [tc ] = month
2 as the engineering hours performed per month are known.
tc =  −3  These depend on the number of engineers working in parallel
4.5114 · ln 10 × hc − 12.102 ∀hc > 105 ; [hc ] = hours
Construction work hours hc can be determined based on
construction costs Cc and construction wage costs cWC .
hc = Cc /cwc (3)
The construction wage costs cwc are determined by dividing
the annual median pay of $50,620 (Bureau of Labor Statistics,
U.S. Department of Labor) by a total of 1790 h actually worked
annually (OECD) and applying that the pay corresponds 69.7%
of the total employer costs (Bureau of Labor Statistics, U.S.
Department of Labor). Applying the 2015 average exchange
rate of 1.1095 $/D (European Central Bank) construction wage
costs of cwc = 37 D /h (rounded) result. The annual median pay
used as basis is “the wage at which half the workers in an
occupation earned more than that amount and half earned
less” (Bureau of Labor Statistics, U.S. Department of Labor).
Compared to the arithmetic mean the median implies the
advantage of being robust against outliers and is applicable
without incorporating additional dispersion measures. Apply-
ing the procedure to engineering wage costs, the result is
cwe = 70 D /h (rounded), based on an annual median pay of
$97,360 (Bureau of Labor Statistics, U.S. Department of Labor).
Construction costs Cc are considered as all labor costs for
workshop and field work that are assigned to assembly and
erection of the production plant. As explained in chapter 1
we use data from a previous work (Sievers et al., 2017) from
which construction costs Cc can be calculated. Details about
the origin of construction costs Cc are given in chapter 4.
The project execution phase tPE is a sum of the construction
phase duration tc and the engineering lead time teLT . As an esti-
mation from analysis of different project schedules, Navarette
(2009) found that tc makes up 80% of the sum of tc and teLT but
has a minimum duration of three months. Applying that, teLT
can be calculated.
teLT = max [0.25 · tc ; 3] [teLT ] = month (4)
Applying this estimation, the duration of the project exe-
cution phase tPE can be calculated.
tPE = teLT + tc (5)
Next the duration of the pre-project phase (tPP ) is cal-
culated. This can be done based on engineering costs Ce ,
which can be taken as a result from a previous work (Sievers
et al., 2017) similar to construction costs, as explained in the
Appendix A. Engineering costs are considered as labor costs
for basic and detail engineering, project management and site
supervision during construction phase and plant start-up.
Similar to calculating construction phase duration, engi-
neering work hours he are determined based on engineering
costs Ce and engineering wage costs cwe (the value of cwe is
70 D /h, as given above).
he = Ce /cwe (6)
From total number of engineering hours he usually a frac-
tion of approx. 12% is allocated to the pre-project phase
(Navarette, 2009). The duration of tPP can be calculated as soon
in this phase (NePP ) and the effective monthly working hours
(2) of an engineer (h = 1790/12 = 149 h/month (OECD)).
M
0.12 · he
tPP = (7)
NePP · hM
Obviously, NePP has to be determined to calculate tPP .
According to the tasks to be performed the number of engi-
 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Fig. 4 – Correlation between construction work hours and field duration.
Data taken from Navarette (2009).
neers changes throughout project execution. Sattler and
Kasper (2000) gives corresponding data about the typical
number of engineers throughout the course of an invest-
ment project. He shows that the number of engineers in the
pre-project phase is rather constant and corresponds to the
number of engineers at the beginning of the engineering lead
time phase. This is plausible as at the end of the pre-project
phase and at the beginning of the engineering lead time phase
basic engineering activities of a similar amount take place
(Sattler and Kasper, 2000). During the engineering lead time
the number of engineers increases gradually. The number of
engineers at the beginning of the engineering lead time cor-
responds to 40.4% of the number of engineers averaged over
the engineering lead time (Sattler and Kasper, 2000). In the
project start phase between the pre-project phase and the
engineering lead time phase (compare Fig. 3) organizational
and economy-based tasks are performed, so no impact on the
number of engineers is observed (Sattler and Kasper, 2000).
Concluding, we assume that NePP can be described in depen-
dency of the number of engineers NeLTA averaged over the
engineering lead time (see Eq. (8)), which will be calculated
in the next step.
NePP = 0.404 · NeLTA (8)
Navarette (2009) found that the average number of engi-
neering hours raised in the engineering lead time corresponds
to 35% of the total engineering hours (he ). Applying this, NeLTA
can be calculated using the result for the calculated engineer-
ing lead time (teLT , for procedure see Eq. (4)).
0.35 · he
NeLTA = (9)
teLT · hM
Applying Eq. (7) the pre-project phase duration tPP can
be calculated. We assume that the staffing and structure of
companies which provide engineering activities will be inde-
pendent of the plant structure (modular or conventional) to
be considered. Based on that assumption we conclude that the
number of engineers working on the investment project would
not be affected and therefore would not differ between plan-
ning modular and non-modular plants. As described above,
the lead time estimation method is meant to be applicable
for both modular and non-modular design. For non-modular
plants, a minimum duration of this phase of three months is
proposed and an upper limit of four months (Navarette, 2009).
By keeping the number of engineers constant, a reduced pre-
101
project phase duration tPP would result for modular plants
compared to non-modular plants, if engineering costs and
therewith he were lower.
To calculate the lead time duration tLT the start-up phase
(tSU ) and the project start phase (tPS ) need to be calculated (cf.
Eq. (1)). For the project start phase we assume that there is
no difference in phase duration between modular and non-
modular plant design as the tasks assigned to this phase
(compare chapter 2.2) are depending on the company’s organi-
zational structure rather than on the technology chosen. The
project start phase duration was estimated with 4.5 months
according to the time for sequentially completing following
main tasks (Navarette, 2009):
• Form project team (0.5 month)
• Select contractor (1.5 month)
• Prepare AFE estimate (0.5 month)
• Wait for AFE approval (2 month)
Finally, the duration of start-up needs to be considered.
It depends on the training of the responsible team and on
smooth operation of the plant. Moreover, this phase strongly
depends on the conditions and complexity of the process to
be performed. Due to these reasons we do consider the dura-
tion of the start-up phase tSU as a cost-independent, fix value
and assume a duration of 1 month as proposed in literature
(Navarette, 2009; Sattler and Kasper, 2000).
4. Results
Input variables of the proposed method are construction
and engineering costs. The approach presented provides the
opportunity to convert these costs into corresponding periods
and thus to calculate an estimated lead time.
The necessary cost information is taken from a previous
work where we propose a method capable of taking the impact
of modular design on FCI into consideration (Sievers et al.,
2017). The FCI estimation bases on a cost factor method pro-
posed by Peters et al. (2003) developed for conventional plants
and transfers the “delivered purchased equipment costs” of
the plant components into FCI using cost factors. The cost fac-
tors used were adjusted to consider different site conditions.
This was necessary to take into account that construction site
conditions change for plant erection with step-wise imple-
mentation of a modular plant. For example, while the first
module is installed under greenfield conditions, the construc-
tion site is already prepared for follow-up modules, which has
102 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Fig. 5 – Resulting cost data according to Sievers et al. (2017) for different values of delivered purchased equipment costs,
used as input parameters for further lead time consideration.
an impact on costs. In the FCI estimation procedure, engineer-
ing and construction costs are calculated separately for the
backbone facility, the first module and follow-up modules. The
cost data generated by that method can be used in the lead
time estimation method so that not only a cost but also a lead
time estimation can be derived for each individual stage of the
modular plant. This is shown in an example below (scenarios
2 and 3).
Moreover, in the FCI estimation method non-scaling costs
are considered that are mostly independent on the plant’s
capacity. This has to be taken into account when considering
containerized production lines limited in scale. For example,
the size and corresponding planning effort of a concentra-
tion measurement is nearly independent on the stream that
is monitored. Additionally, there is a base effort of engineer-
ing (e.g. project team assembly, contractor selection, start-up
supervision, preparation of approval for expenditure) which is
not linearly dependent on equipment costs (Navarette, 2009).
Like for costs this fact influences lead time, accordingly.
To achieve the same production capacity like a conven-
tional plant a modular plant requires a larger number of
individual equipment items when these are limited in size
(Bramsiepe, 2014). This leads to an increased complexity for
engineering and construction and thus increased costs con-
sidered in the FCI estimation method and corresponding
impact on lead times.
On the other hand, there are changes induced by mod-
ularization that lead to a reduction of engineering and
construction effort. For a modular plant setup construction
cost reduction is expected by using standardized compo-
nents and installation procedures in the modular plant (Lier
et al., 2015; Behr et al., 2003). Furthermore, construction can
take place under workshop conditions and less field labor is
required for module assembly (Lier et al., 2015), which was
considered in the FCI estimation method (Sievers et al., 2017).
Engineering costs are expected to be reduced using a modular
approach due to pre-designed structures and design repetition
(Buchholz, 2010; Lier et al., 2015), which was also investigated
and considered in the FCI estimation method (Sievers et al.,
2017) and takes effect for lead time estimation.
Sievers et al. (2017) demonstrated the FCI estimation
method using a brief example of a modular production plant
that includes three identical modular production lines and a
non-modular backbone facility that provides energy and util-
ity supply. The process used is a generic example that is based
on a reference process and a modular process both developed
as an example for small scale specialty chemicals production
within the F3 factory project in 2013. The delivered purchased
equipment costs of the continuously operated reference plant
with one production line are 2 million D , the number of main
equipment items is 20 and the scale-independent “fixed
equipment costs” are 780,000 D . Applying the rule of economy
of scale (degression exponent = 0.7), a single production line
with a third of production capacity of the continuously oper-
ated reference plant amounts for 927,000 D . Both plants are
built under greenfield conditions (Sievers et al., 2017).
The results of applying the FCI estimation method were
used for calculating lead times and comparison between con-
ventional and modular plant design. The resulting data for
construction (Cc ) and engineering costs (Ce ) used for lead time
estimation can be extracted from the FCI calculation method
using the procedure explained in the Appendix A.
The resulting cost data (construction and engineering costs
as well as fixed capital investment for both, conventional and
modular construction) that is the basis for lead time calcu-
lation for the example chosen are shown in Fig. 5. To cover
a range of investment sizes and to evaluate the resulting lead
times the FCI calculation was done for different plant sizes and
resulting values of delivered purchased equipment costs rang-
ing from the given initial value up to 20 million D . Thereby the
number of equipment items and fixed equipment costs were
kept constant, no numbering up was considered.
Based on the cost data given the lead time calculation pro-
cedure presented in chapter 3 was applied. The resulting lead
times are shown in Fig. 6. In this “scenario 1”, resulting lead
times for the conventional plant range from 16.5 to 32 months
for investment sizes (FCI) of 3 to 95 million D , which is a good
agreement with literature data (Rauch, 1998; Brodhagen et al.,
2012).
Resulting lead times for corresponding modular plants are
13.9–26.4 months. This corresponds to a lead time reduction
of 15.5%–17.3%. Whereas in the case of small investment the
reduction of pre-project phase is determining for lead time
reduction (1.5 vs. 3 months), for large investments the major
leverage comes from reducing the project execution phase
 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Fig. 6 – Scenario 1 — resulting lead times and project phase durations for conventional and modular construction of
different sizes of the investment project.
(18.6 vs. 22.6 months), due to a significant reduction in con-
struction effort.
For this result a scale-up of each single equipment item
was assumed for different production capacities, leading
to increasing delivered purchased equipment costs without
changing the actual setup of the plant. The idea of modular
production plants of course is to use segmented construction
elements to gain flexibility concerning capacity and product
variability, as described in chapter 1. Regarding the lead time,
the construction elements to be distinguished in the F3 fac-
tory approach taken as example for modularization are the
backbone facility and the production modules. In the FCI cal-
culation methodology used as starting point for lead time
estimation learning effects originating from module design
repetition are considered (Sievers et al., 2017). Therefore, the
first production module and follow-up modules are distin-
guished as well.
Hence, two further scenarios should be evaluated. In sce-
nario 2, the size of the production modules and the backbone
facility are scaled as before but a breakdown of lead time pro-
portions of each individual construction element is used to
determine which element adds which portion to the total lead
time. In scenario 3 the module size is kept constant but the
number of modules is varied according to the plant capac-
ity. In this case the size of the backbone facility again scales
with the plant capacity as it needs to be able to host the total
number of production modules.
Fig. 7 (scenario 2) shows resulting lead times separately for
the backbone facility, the first production module and follow
up modules for different plant sizes. The number of mod-
ules was fixed to three, which means the backbone facility is
constructed to host a maximum of three modules. Assuming
parallel planning and construction of the production modules
and the backbone facility the longest lead time determines the
total lead time which in this example is the lead time of the
backbone facility and ranges between 13.7 and 23.3 months.
Compared to the conventional plant this corresponds to a lead
time reduction between 2.8 and 8.7 months.
According to the FCI calculation procedure (Sievers et al.,
2017) the backbone facility is not considered as a modular
unit but as a conventionally built construction in which the
modular production lines are hosted. As a consequence, the
construction and engineering costs of the backbone facility
are directly correlated with the production capacity and the
lead time accordingly. For the production modules cost reduc-
tions from modularization take effect resulting in a reduction
103
of the project execution phase. Moreover, for modular plant
lead time calculation the available number of engineers cor-
responds to the number of engineers of the conventional plant
it is compared with (see NeLTA , chapter 3), leading to a reduc-
tion of the pre-project phase. As a result, the lead time of
the production modules is essentially determined by the cost-
independent periods considered in lead time calculation and
do rarely correlate with the plant capacity in the range inves-
tigated. Due to the same reason, the difference between the
first production module and follow-up modules is negligible.
In scenario 3 a fixed production module size is assumed.
Different plant capacities are represented by a corresponding
number of identical production modules. This approach would
allow preproduction of modules and thus a significant reduc-
tion of lead time regarding capacity expansion. Nevertheless,
to compare such an approach with a conventional plant setup
the same capacity needs to be considered, which is the design
capacity of the conventional plant and the design capacity
of the final expansion stage of the modular plant. According
to the FCI calculation approach the backbone facility is not
considered as a modular setup (Sievers et al., 2017). Thus it
needs to be designed to host the number of production mod-
ules according to the final expansion stage of the complete
modular plant.
It is expected that a backbone facility designed to host a
higher number of production modules is more complex than a
facility hosting a lower number of production capacity offering
an identical total capacity. Hence this should also be reflected
in costs and lead time of the backbone facility, which would
diminish the advantages expected of the approach of a mod-
ular plant.
In Fig. 8 the lead time calculation results for scenario 3 are
shown. The grey dots show the maximum number of produc-
tion modules hosted by the corresponding backbone facility.
The resulting delivered purchased equipment costs of the pro-
duction modules amounts to 0.5 million D per module. Due
to the higher complexity and costs of the backbone facility
the lead time increases compared to scenario 2. It amounts
to 12.6–28.7 months. Assuming parallel processing of all con-
struction elements the lead time of the complete modular
plant is determined by the backbone facility’s lead time as
it needs to be finished before start-up of the modular plant.
Compared to the conventional plant this leaves a lead time
saving potential between 2.6 and 3.9 months, which is less
than in scenario 2.
104 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Fig. 7 – Scenario 2 — lead times for conventional and modular construction of different plant sizes represented by delivered
purchased equipment costs.
Fig. 8 – Scenario 3 — lead times for conventional and modular construction of different sizes of the investment project with
a backbone facility that needs to host a maximum number of modules that corresponds to the capacity of the conventional
construction (grey dots).
This result shows a demand to decrease the lead time
of the backbone facility as its reduction offers the largest
lever to decrease total lead time for the modular plant.
Such a reduction could be achieved e.g. by modularization of
the backbone facility. Assuming that the backbone facility’s
capability for hosting production modules could be adapted
stepwise according to the desired number of production mod-
ules to be hosted, the total lead time of the plant might no
longer be determined by the backbone facility’s lead time and
could be reduced significantly. If the lead time of such mod-
ular backbone facility was comparable to the modules’ lead
time (12.6 months), depending on the plant’s capacity total
lead time reduction compared to the conventional plant would
be between 3.9 and 18.7 months (compare Fig. 8) or 23%–60%,
representing a very significant reduction. From this result we
can conclude that an approach for modular plant setup like
the F3 factory approach taken as example needs to include
the modularization of the backbone facility as well. Alter-
natively, a non-modular but standardized backbone facility
design could be useful. This way engineering and construction
of the backbone facility could be done once, independently
from the process to be performed or specifically for a class of
processes, which would allow to reduce lead time especially if
a numbering up approach for the backbone facility itself would
also be considered.
5. Conclusion
A method for calculating lead times of modularly designed
plants is presented. Phases that have to be considered for lead
time estimation were identified. Using the method presented,
the phase durations and the overall lead time can be calcu-
lated using the expenses for engineering activities and the
construction work to be conducted in an investment project
as starting parameters. To convert expenses for construction
work and engineering activities into time intervals correla-
tions for project staffing and salary information were used.
Certainly, the parameter values used are worth being dis-
cussed in detail, but we expect them to be in the right order.
Since the method is generic the factors may be adjusted when
new knowledge is available.
The proposed method can be applied to non-modular
conventional plant design and modular plant design, which
enables comparative studies. For such comparison the cost
basis provided by Sievers et al. (2017) for modular and non-
modular plant design was used, which considers the impact
of modular design on construction and engineering costs.
The approach of modularization used as an example was
also used in the F3 factory project. By that approach, the mod-
ular plant is composed of modular production lines which are
housed by a non-modular backbone facility. Because the back-
bone facility is non-modular, it has to be designed according
to the final (maximum) capacity of the production plant and
cost reductions from modularizations cannot take effect.
As a consequence, the backbone facility adds the biggest
proportion to the lead time. Compared with a conventional
plant total lead time reduction is between 2.6 and 5.5 month
depending on the size of the investment. If the lead time of the
backbone facility would be significantly shorter, total lead time
of the modular plant design could be reduced substantially
assuming parallel assembly of the modules and the backbone
facility.
 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106 105

For that the backbone facility would need to follow a stan- fraction, ci for the construction cost fraction of the correspond-
dardized design or ideally a modular design as well so that a ing cost item (columns three and four, Table 1).
step-wise expansion of the backbone facility would be pos-

sible. By that the investment into the first production line CMat\OVC = ai · mi where i = COCW , COB , . . ., CAEI , FFEC (10)
providing initial production capacity would represent a com- i
parably small investment with a short lead time. If the lead
time of such modular backbone facility was comparable to 
CConstr\OVC = ai · ci (11)
the production lines’ lead time, depending on investment
i
size total lead time reduction would be between 3.9 and 18.7
months representing a very significant reduction of 23%–60% Then the material cost fraction (mOVC ) and the construction
compared to the lead time of the conventionally designed ref- cost fraction (cOVC ) of the contingency costs can be calculated:
erence plant. This is considered as the maximum lead time
reduction that can be achieved through modular plant design CMat\OVC
and would represent a major potential for speeding up con- mOVC = (12)
CMat\OVC + CConstr\OVC
struction of chemical plants.
The results may be validated when further data gets avail- cOVC = 1 − mOVC (13)
able, e.g. by future construction of plants according to the
modular F3 factory design. However, the identified lead time Finally, the sum of construction costs CC can be deter-
reductions would substantially improve the economy by gen- mined:
erating earlier sales resulting in a shorter pay-back period.
This in turn helps reducing the investment risk and allows CC = COVC · cOVC + CConstr\OVC (13)
going for opportunities that otherwise were too risky. Ulti-
mately this creates an increased degree of business flexibility.
The engineering costs Ce are calculated as a separate cost
item COVE in (Sievers et al., 2017)
Acknowledgment
Ce = COVE (15)
The research leading to these results has received funding
from the European Community’s Seventh Framework Pro-
References
gramme (FP7/2007-2013) under grant agreement no. 228867.

Appendix A.
In the following, the calculation of the construction costs
Cc and engineering costs Ce based on data generated in the
established FCI calculation procedure (Sievers et al., 2017) is
explained. Basis for calculating Cc and Ce are the single cost
items shown in Table 1, which are considered in the FCI esti-
mation method described in a previous work (Sievers et al.,
2017).
First, using the calculation procedure described by Sievers
et al. (2017) the values of the single cost items (listed in the
Table 1) and – accordingly – the sum of material costs (CMat \OVC )
and construction costs (CConstr \OVC ) without contingency costs
(COVC ) can be determined. In the following equations, ai stands
for the costs of each cost item (COCW , COB ,. . .) listed (see first
and second column, Table 1), mi stands for the material cost
Baerns, M., Behr, A., Brehm, A., Gmehling, J., Hofmann, H.,
Onken, U., Renken, A., 2006. Technische Chemie. Wiley-VCH
Verlag GmbH & Co. KGaA, Weinheim(DE).
Behr, A., Brehme, V.A., Ewers, C.L.J., Grön, H., Kimmel, T.,
Küppers, S., Symietz, I., 2003. Neue Entwicklung bei Anlagen
und Verfahren för die Produktion von Pharmawirkstoffen.
Chem. Ing. Tech. 75 (4), 417–427.
Bernecker, G., 2001. Planung und Bau verfahrenstechnischer
Anlagen: Projektmanagement und Fachplanungsfunktionen.
Springer Verlag.
Bramsiepe, C., Kussi, J., Schembecker, G., 2011. Projektierung von
Chemieanlagen: Mit der 50%-Idee vom Produkt zur
Produktionsanlage. Chem. Tech. 40, 50–52.
Bramsiepe, C., Sievers, S., Seifert, T., Stefanidis, G., Vlachos, D.,
Schnitzer, H., Muster, B., Brunner, C., Sanders, J., Bruins, M.,
Schembecker, G., 2012. Low-cost small scale processing
technologies for production applications in various
environments — mass produced factories. Chem. Eng.
Process. Process Intensif. 51, 32–52.

Table 1 – Extract from results elaborated in a previous work, giving an overview about the cost items considered (Sievers
et al., 2017).
Cost item ai Material costs fraction mi Construction costs fraction ci

Outside battery limits construction work COCW 60.4% 39.6%

Outside battery limits buildings COB 58.8% 41.2%
Outside battery limits facilities COF 58.8% 41.2%
Delivered purchased equipment costs BdPEC 100% 0%
Piping material CCPM 100% 0%
Material for electrical installations CCME 100% 0%
Supports, platforms, etc. CCSP 50% 50%
Equipment insulation and painting CCIP 36.3% 63.7%
Inside battery limits buildings, HVAC CCIB 50.4% 49.6%
Auxiliary work CAAW 0% 100%
Equipment assembly CAEA 0% 100%
Piping assembly CAPA 0% 100%
Electrical installation CAEI 0% 100%
Instrumentation BFEC 75% 25%
106 chemical engineering research and design 1 2 8 ( 2 0 1 7 ) 96–106
Bramsiepe, 2014. Information technologies for innovative process
and plant design. Chem. Ing. Tech. 86 (7), 966–981.
Brodhagen, A., Grünewald, M., Kleiner, M., Lier, S., 2012.
Increasing profitability by accelerated product- and process
development with modular and scalable apparatuses. Chem.
Ing. Tech. 84 (5), 624.
Buchholz, S., 2010. Future manufacturing approaches in the
chemical and pharmaceutical industry. Chem. Eng. Process.
Process Intensif. 49, 993––995.
Buchholz, S., 2013. Bayer Technology Services GmbH (project
coordinator), Final report F3 Factory project. F3 Factory project,
http://www.f3factory.com/scripts/pages/en/newsevents/F3
Factory final report to EC.pdf. (Accessed 30 January 2016).
Burdorf, A., Kampczyk, B., Lederhose, M., Schmidt-Traub, H.,
2004. CAPD — computer-aided plant design. Comput. Chem.
Eng. 28 (1–2), 73–81.
Bureau of Labor Statistics, U.S. Department of Labor.
Occupational Outlook Handbook, 2016-17 Edition, Plumbers,
Pipefitters, and Steamfitters, on the Internet at
http://www.bls.gov/ooh/construction-and-extraction/plumbers
-pipefitters-and-steamfitters.htm. (Visited 15 October 2016).
Bureau of Labor Statistics, U.S. Department of Labor. Employer
Costs For Employee Compensation — June 2016, Private
industry workers,
http://www.bls.gov/news.release/pdf/ecec.pdf. (Visited 15
October 2016).
Bureau of Labor Statistics, U.S. Department of Labor.
Occupational Outlook Handbook, 2016-17 Edition, Chemical
Engineers, on the Internet at
http://www.bls.gov/ooh/architecture-and-engineering/
chemical-engineers.htm. (Visited 15 October 2016).
European Central Bank. Euro exchange rates USD, Average
2015-01-01 to 2015-12-31.
https://www.ecb.europa.eu/stats/exchange/
eurofxref/html/eurofxref-graph-usd.en.html. (Data extracted
on 16 October 2016).
Festel, G., Hassan, A., Leker, J., 2001. Betriebswirtschaftslehre für
Chemiker: eine praxisorientierte Einführung. Springer Verlag.
Grundemann, L., Schoenitz, M., Scholl, S., 2012. Shorter
time-to-market with micro-conti processes. Chem. Ing. Tech.
84 (5), 685.
Helmus, F.P., 2008. Process Plant Design. Wiley VCH, Weinheim.
Jameson, P.H., 2007. Is modularization right for your project?
Hydrocarbon Process. (December), 67–71.
Kampczyk, B., Hicking, B., Schmidt-Traub, H., 2004. More efficient
plant design by modularization? Chem. Eng. Technol. 27 (5),
481––484.
Khosrowshahi, F., 2002. Project cost-time optimal solutions. In:
Greenwood, D. (Ed.). 18th Annual ARCOM Conference, 2–4
September 2002, University of Northumbria. Association of
Researchers in Construction Management, Vol. 2, 839–848.
Löb, P., 2013. EU FP7 Project CoPIRIDE — towards new production
and factory concepts for a sustainable and competitive
European chemical industry. Green Process. Synth. 2 (5),
379–380.
Lier, S., Wörsdörfer, D., Grünewald, M., 2015. Wandlungsfähige
Produktionskonzepte: Flexibel, Mobil, Dezentral, Modular,
Beschleunigt. Chem. Ing. Tech. 87 (9), 1147––1158.
Madauss, B., 2000. Handbuch Projektmanagement.
Schaffer-Poeschel, Stuttgart.
Mosberger, E., 2012. Chemical plant design and construction. In:
Ullmann’s Encyclopedia of Industrial Chemistry. Wiley-VCH,
Weinheim.
Organization for economic co-operation and development
(OECD). Average annual hours actually worked per worker
(basis: total employment).
https://stats.oecd.org/Index.aspx?DataSetCode=ANHRS. (Data
extracted at 16 October 2016).
Navarette, Pablo F., 2009. Planning, Estimating, and Control of
Chemical Construction Projects.
Peters, M.S., Timmerhaus, K.D., West, R.E., 2003. Plant Design and
Economics for Chemical Engineers, 5th edition. The
McGraw-Hill Companies, Inc.
Rauch, J., 1998. Mehrproduktanlagen. Wiley-VCH, Weinheim.
Sattler, K., Kasper, W., 2000. Verfahrenstechnische Anlagen:
Planung, Bau und Betrieb. Wiley-VCH Verlag GmbH & Co.
KGaA, ISBN 9-783-52728459-7.
Schmidt-Traub, H., Holtkötter, T., Lederhose, M., Leuders, P., 1999.
An approach to plant layout optimization. Chem. Eng.
Technol. 22 (2), 105––109.
Seifert, T., Sievers, S., Bramsiepe, C., Schembecker, G., 2012. Small
scale, modular and continuous: a new approach in plant
design. Chem. Eng. Process. Process Intensif. 52, 140–150.
Seifert, T., Fakner, P., Sievers, S., Stenger, F., Hamers, B., Priske, M.,
Becker, M., Franke, R., Schembecker, G., Bramsiepe, C., 2014.
Intensified hydroformylation as an example for flexible
intermediates production. Chem. Eng. Process. Process
Intensif. 85, 1–9.
Sievers, S., Seifert, T., Franzen, M., Schembecker, G., Bramsiepe,
C., 2017. Fixed capital investment estimation for modular
production plants. Chem. Eng. Sci. 158, 395–410.
Wagner, W., 2003. Planung im Anlagenbau, 2nd edition. Vogel
Business Media/VM.