Project management operations and building performance in the construction industry: A multi
method approach of applied in a UK public office building
Papachristos, G., Jain, J., Burman, E., Zimmerman, N., Mumovic, D., Davies, M.
Abstract
The “performance gap” in the UK building industry is a persistent problem as new building
development projects underperform more often than not. Underperformance has to be addressed as
the building sector is responsible for a large share of CO2 emissions in the UK. The “performance
gap” arises in part, because building project development involves operations with several stages
and actors of different motivations. The outcome of the building project in terms of quality is
important and has implications for energy consumption, carbon emissions and occupant well-being.
We develop a system dynamics model of building project development operations to explore
building quality implications for energy consumption and Indoor Environmental Quality (IEQ). To
do this, we couple the system dynamics model to a building physics model and apply them in an
empirical case of a recently completed building project. The building performance model is
developed and calibrated to reproduce the actual energy performance of the building based on one
year of monitoring and post commission data. This is used as a reference point for the system
dynamics model that explores additional scenarios of how project operations could deliver better
total building performance.
Keywords: project management, operations, simulation, low carbon, system dynamics
1 Introduction
The building sector, which includes residential and commercial structures, accounts for almost 21%
of the world’s delivered energy consumption in 2015 (EIA, 2017). In the EU, buildings are
responsible for 40% of energy consumption and 36% of CO: emissions!. For example, the
residential sector in UK accounted for 18% of all CO2 emissions in 2016 (DBEIS, 2017), and the
building sector accounts for more than 45% of UK emissions (Oreszczyn and Lowe, 2010). It is
estimated that energy efficiency strategies can reduce a building’s energy consumption by 50% to
70% (Zervos et al., 2010). Urgent and ambitious measures are required for the adoption of state-of-
the-art performance standards in new and retrofit buildings (IPCC, 2014).
The UK government in 2009 adopted an 80% target of total emissions reduction by 2050.
This would require faster emission reduction in the building sector than the current rate (Oreszczyn
and Lowe, 2010). Reductions in building energy consumption must also not generate unintended
consequences in terms of indoor environmental quality (IEQ) and other performance metrics
(Davies and Oreszczyn, 2012; Shrubsole et al., 2014; Shrubsole et al., 2018). Achieving the CO2
reduction targets by 2050 cannot just depend on combinations of technologies that have dominated
over the last three decades or simply a continuation of the current trends (Lowe, 2007).
This poses a considerable challenge as behavioural and factors specific to construction supply
chain (CSC) partner interactions in building design, construction and operation project stages
influence the long term building quality, energy consumption, and IEQ (Bendoly and Swink, 2007;
O’Brien et al., 2009; Alencastro et al., 2018; Gram-Hanssen and Georg, 2018). In this respect, UK
1 https://ec.europa.eu/energy/en/topics/energy-efficiency/buildings (accessed 13/2/2018)
government reports have highlighted the need for improvements in the historically fragmented UK
building industry (Latham, 1994; Egan, 1998). Project improvements could be achieved through
greater integration, and operation coordination at the organisational level between clients and
suppliers (Turner and Miiller, 2003). Since the publication of the reports, supply chain collaboration
has increased in UK construction industry operations practices (Meng, 2013). Despite, some
improvement in the energy performance of the existing non-domestic stock, performance gaps
remain between the intended and actual performance of new and refurbished buildings” (Cohen et
al., 2001; De Wilde, 2014).
One reason for this, is the complex and ineffective UK regulatory landscape of incentives for
energy efficiency in commercial buildings. This is compounded by the lack of focus by all partners
involved in a CSC about what works in practice when it comes to reductions of building energy use
and emissions, and what works in practice (Cohen and Bordass, 2015). UK policy should focus
more on actual energy use than theoretical estimates, and behavioural drivers for improvement as
they are at least as important as financial ones (Cohen and Bordass, 2015) *. Given the growing
need to achieve low carbon emissions in all industrial sectors by 2050, is it possible to achieve
further building performance improvements through collaboration in CSC operations?
The focus on physical project work flows must be complemented with a focus on inter-stage
collaboration between CSC project partners to account for UK industry fragmentation. Supply chain
collaboration has certain precedents: the goals alignment of project partners and client, the trust
between them, information sharing, and antecedents: the delivery of value to the client (Bendoly
and Swink, 2007; Hanson et al., 2011; Wong et al., 2012). However, the implications of these
antecedents on building performance are not explored in recent project operations management
modelling and simulation work (Rahmandad and Hu, 2010; Han et al., 2013; Parvan et al., 2015).
The current paper tries to address this gap, explore and document potential solutions to this
problem (Holmstrom et al., 2009). This is done by means of a modelling framework that seeks a
sense of theoretical generality while being situationally grounded, methodologically rigorous and
practically relevant (Ketokivi and Choi, 2014). The project management part of the framework is
sufficiently generic and the building physics part is used to ground the framework in a particular
context. The framework aims to explore the effect of CSC collaboration and operations
management on operational building performance and IEQ on a case by case basis.
It is the first attempt to bridge buildings and performance gap. The paper follows a multi-
methodology approach that combines the technical and social aspects of project management
(Mingers and Brocklesby, 1997). Two simulation methods from different domains of expertise are
combined in a novel way. System dynamics is used for project management and supply chain
collaboration modelling (Sterman, 2000; Lyneis and Ford, 2007; Mingers and White, 2010), and
building physics modelling for building performance (Hensen and Lamberts, 2011). System
dynamics is often combined with other methods (Howick and Ackerman, 2011; Zolfagharian et al.,
2018) and the framework development is geared to tackle a class of problems rather than a single
case (Forrester, 1961). System dynamics modelling and simulation has been proposed and explored
as a complementary methodological tool to low carbon transition case study research (Papachristos,
? Committee on Climate Change (2014). Meeting carbon budgets — 2014 Progress york to parliament. London.
http://www. theccc.org.uk/publication/meeting-carbon-budgets-2014-pro to-parliament/
3 Committee on Climate Change (2014). Meeting carbon budgets — 2014 progress Bee to parliament. London.
http://www. theccc.org.uk/publication/meeting-carbon-budgets-2014-pro rt-to-parliament/
2012, 2014a, b; Holtz et al., 2015; Papachristos and Adamides, 2016; Papachristos, 2017; Kéhler et
al., 2018; Papachristos, 2018).
The framework is applied to a recently completed public office building in UK that followed a
design and build procurement approach (Molenaar et al., 1999). Through seven hour-long, semi
structured interviews, and a workshop with project stakeholders that focused on the particular
project and industry related issues, it is clear that the case has a number of characteristics that make
it an appropriate choice for research (Y in, 2003): (i) the building energy performance target set in
consultation with the client was Display Energy Certificate A (DEC), placing it in the top 15% in
terms of performance in UK, (ii) the project followed a soft landings approach which aims to keep
designers and constructors involved in the performance of buildings beyond completion (De Wilde,
2014), (iii) partner alignment and commitment was high as they considered it to be a flagship
project in terms of energy and IEQ performance.
The rest of the paper is structured as follows. Section 2 presents an overview of the case
building. Section 3 provides the conceptual foundation for the framework. Section 4 present the
system dynamics model and discusses how it is coupled to the building performance model. Section
5 presents result of the study for the building and explores the effect of project operations factors.
Section 6 discusses limitations and concludes the paper.
2 The Building Case
The case concerns a public office building complex in the UK with 4 storeys, designed for 450
staff. The building is intended for long term use and the client has a vested interest to achieve low
operational use costs. The target of the project is to achieve a Display Energy Certificate (DEC) A
rating for building performance. The novelty is that the DEC A goals is written into the contract
(but no IEQ goal). The project is the first to employ a four-year, post commission, “soft landing”
approach‘ during which designers and constructors will try to improve building energy efficiency
(De Wilde, 2014). The building is close to but has not yet reached DEC A performance, three years
after its commission. Nevertheless, it has won several industry awards as an exemplar for UK
public buildings and received wide publicity with a lot of sustainability themed tours around the
building attended by industry professionals.
Interviews with seven industry experts that were stakeholders in the research project, were
conducted by the same researcher to ensure consistency. They confirmed that such a strong client
emphasis on building energy performance is still a niche market segment. Five of them were
directly involved in the project and participated in a focused workshop. This provided the research
team the opportunity to juxtapose the content of their interviews with the retrospective discussion
about aspects of the project. A consensus view formed around some of the points raised in the
workshop.
3 The Modelling Framework
The framework adopts a flow view of production in construction supply chains (CSC) (Vrijhoef and
Koskela, 2000). The core logic of project management model draws on prior system dynamics work
(Ford and Sterman, 1998; Parvan et al., 2015). It involves workflows of project tasks completion,
defects° that arise in the process, and the decision logic that drives these flows within project and
4 https://www.bsria.co.uk
5 Semantics note: tasks and defects are standard terms in the system dynamics project management literature. Defects
lead to a deviation in project performance. In the building science literature deviation from project performance arises
between project stages and contribute to building quality. The logic generates project parmer
collaboration dynamics.
The framework uses Case Project Input (Figure 1) on building project characteristics: project
timing, resources, stages, and organizational aspects, and the building performance gap i.e. the
building areas where known operational building performance deviates from design targets, a
widely applied definition in the UK (Cohen et al., 2001). The areas and gap magnitude are
established through a Building Performance Model. The system dynamics (SD) model is calibrated
and uses the Case Project Input to endogenously generate Building Quality Indices that correspond
to the building areas where known operational Total Building Performance deviates from its design
targets. The underlying assumption in coupling the two models is that building quality can be used
as a proxy for building performance (Alencastro et al., 2018). The SD model is then used to explore
the operational options that could result in better building quality and thus Total Building
Performance i.e. energy consumption and IEQ.
Case Project Inputs
¥
SD Project Management Model
(2ariner Interaction: Alignment )
(Decisions: CSC Project Partners )
(Ho ws: Information, Tasks, Defects)
¥.
Building Quality Indices
¥
Building Performance Model
Building Total Building
Characteristics Performance
Figure 1 The modelling framework combining project management and building performance
3.1 The Construction Supply Chain
The project management model is based on a simplified construction supply chain (CSC) (Love et
al., 2004). Individual organizational actors are aggregated to the organizational level, and CSC
organizations are aggregated to the stage level. This aggregation is possible as social entities in
hierarchies above the level of individuals can form parts of social mechanisms (Hedstrém and
Swedberg, 1998; Papachristos, 2018). Thus, the CSC consists of design, construction, and
operation-client stages each with a respective aggregate actor teams and a related remit of
responsibilities (Figure 2).
CSC task flows are based on Ford and Sterman (1998). Tasks are subject to Quality Testing at
the end of each stage to find defective tasks that lower building quality. A modification on Ford and
Sterman (1998) is introduced to increase model realism in line with real construction practice. An
additional task flow (solid grey arrows), is used to account for the flow of defective tasks or
workarounds to downstream stages due to time pressure, negligence or other limitations (Morrison,
2015; Aljassmi et al., 2016). Project partners choose to do workarounds rather than engage with
from technical defects, and/or deviation from set value parameters. Acknowledging the difference, the terms defects and
deviation are used interchangeably in the text.
4
upstream stages to find a collaborative solution that requires more coordination and time (Aljassmi
et al., 2016).
Information exchange is taken into account as it is important for supply chain performance
(Lee et al., 1997). Partners exchange information in- and between stages to monitor delivery of
work and quality and guard against opportunism (Turner and Miiller, 2003). Finally, alignment
between actors, facilitates a greater effect as partners share common goals and a shared
understanding of how this can be achieved in the project.
1. Design 2. Construction 3. Operation
2. Construction
T. Design
Team
Building Design ity \_, (Construction)
Brief Work _}*-{ Testing Work}
— > Completed Tasks in Each Stage
Forward Workarounds <
Return Defects for Correction [____] Activities in Stage
Alignment of Actors
-- Information Exchange
Figure 2 Conceptualization of project stage physical flows between design and construction stages
The conceptual CSC is formalized in an SD model. Figure 3 shows a simplified structure for
illustration purposes, of the core task flows between Design (1) and Construction (2) (the operation
stage has the same structure). The flows in grey depend on the particular level of cooperation
between CSC partners. In each stage the work and test activities utilize a simple rework cycle
formulation.
Tasks Retumed Tasks Requiring
Upstream 2 Coordination 2
Tasks Requiring Tasks Released
Coordination 4
Downstream 1
Inter Stage Defect
Discovery Rate 2
Intra Stage Defect inter Stage Detect Intra Stage Defect me
Upstream Tasks Discovery Rate Discovery Rate 1 Discovery Rate 2
Received
Tasks for Quality
Testing 2
Tasks not 5 Tasks For Quality tacq Approval Tasks not
Completed 1 Task Completion Testing 1 Rate 1 Completed 2 Task Completion
Rate 1 Rate 2
Figure 3 Core stock and flow task structure of two stages of the SD project management model
A co-flow structure (Sterman, 2000) accounts for defects in each stage (Figure 4). Tasks and defects
differ and need to be accounted for on a case by case basis as each building is a unique project. An
array in the SD model accounts for building areas, with corresponding tasks and related defects.
The array forms the interface with the building performance model that enables a detailed analysis
of the operational building performance.
Generate Internal
Defects in
Completion 1
Generate Internal
Defects in Initial
Completion 2
Defects Released
1
Undiscovered
Defects 1
efect Undiscovere Defects Released
Discover Internal
Defects 2
Defects Design
Generate Defects lb Stage Generate Defects Discover Internal
in Iteration 4 in Iteration 2 Defects 2
Discover Inter Release Defects
Stage Defects 2 with Work
arounds 1
Known Defects 1 Known Defects 2
Discover Inter
Correcting Correcting Stage Defects 3
Defects 1 Defects 2
Corrected Defects Corrected Defects
1 Ps
Figure 4 Defect stock and flow structure of two stages of the SD project management model
The task and defect flow structures have a decision and control logic that is driven by partner
alignment and information sharing between CSC stages.
4 System Dynamics Model Development
The SD construction project model is developed in Powersim ©® and is based on reviewed
literature, and Ford and Sterman (1998)’. Two of the authors with industry experience provided a
sanity check throughout model development.
4.1 Partner Alignment
Organizational alignment research spans the strategic management, supply chain management and
project management literatures, and links organizational activities with strategy, and competitive
advantage (Powell, 1992; Williams and Samset, 2010; Hanson et al., 2011; Wong et al., 2012;
Samset and V olden, 2016; Adner, 2017). It requires clear cause and effect mechanisms, a consensus
on strategic goals and actions at the operational level and behaviours towards an operational
outcome. Goals provide a rationale for prioritization, resource allocation, and action in project
management settings (Brenner, 1994). Goal alignment arises from the logical structure of the
project, and the causal link from the basic client needs, defined goals, to the delivery of project
results, their outcomes and long-term benefits after the project is terminated.
In the model, intra-stage alignment A; reflects the level of shared goals in stage i. An initial
level of alignment A?, is assumed to exist based on prior collaboration among partners. This level
was elicited from interviews and the workshop (see Appendix C). CSC partners must have and
sustain a minimum level of alignment and coordination to deliver value to their clients (Gattorna,
2009; Williams and Samset, 2010). Alignment is dynamic as partners make sense of a project, work
towards its delivery, and cope with ambiguity, uncertainty and complexity (Weick, 1995). Intra-
stage alignment A; increases with stage duration, which gives partners a chance to interact more. A;
is a stock that accumulates with the rate of aggregate partner engagement E; per month and faces
diminishing returns with stage duration Lj. A; erodes with partner conflict, or as partner
5 The model is developed in Studio 10. The complete list of SD equations is in Appendix A. The SD and building
physics models are available upon request from the authors.
? The detailed working paper version is available from https://dspace.mit.edu/bitstream/handle/1721.1/2644/SW P-3943-
36987273.pdf?sequence=1 (accessed 06/02/2018)
6
participation nears its deadline Dj; and other projects become more pressing. Suppressing time
subscript t for clarity, Ai is given by:
Ai = f, (4? +2 —4) ae (1)
Inter-stage alignment A,; between stage i andj reflects the level of shared goals across project
stages. An initial level of alignment A?; is justified as project partners had a history of prior
collaboration discussed in the workshop. A high level of A,; implies that CSC project partners are
willing to receive and rework defects from downstream stages to improve the overall building
quality. An initial level of alignment A?, is assumed as project partners had collaborated previously,
and they aimed to deliver a high-performance building. It is assumed that intra-stage partner actions
in the project are sufficiently visible and considered in their subsequent reciprocal behaviour, so
that the motives and constraints of project partners should play a significant role as well (Bendoly
and Swink, 2007). Ajjis assumed to increase with A; and A, and is given by:
Ajj = Ay X Aj + AB, (2)
Alignment is important for performance as a precedent for coordination and information sharing, to
eliminate defects, reduce rework and defects, and increase supply chain performance (Briscoe et al.,
2004; Kache and Seuring, 2014; Alencastro et al., 2018). High coordination generally results in
high quality teamwork and has a positive influence on multi-partner project performance, problem
solving, and dispute handling (Hoegl and Gemuenden, 2001; Dietrich et al., 2010; Baiden and
Price, 2011; Suprapto et al., 2015).
Information sharing is an important moderator of coordination and shared understanding of
project dynamics and project performance (Cohen and Bailey, 1997; Bendoly, 2014). Information is
required to complete tasks and reduce uncertainty related to them when they cannot be pre-planned
(Galbraith, 1973; Tushman and Nadler, 1978). Information facilitates transparency between CSC
partners, responsiveness and lower uncertainty, collaborative planning and risk management
(Frohlich and Westbrook, 2002; Barratt, 2004; Soosay et al., 2008; Wong et al., 2012). Project
partners share information to coordinate their activities, handle operational and technical issues, and
deliver client value (Jingmond and A gren, 2015).
Failure to appreciate the criticality of information flows among stages, and the upstream and
downstream effects they can have, can lower communication levels, information quality and
increase project rework (Love et al., 2008; Tribelsky and Sacks, 2010; Jingmond and Agren, 2015).
The quantity of rework in design and construction stages is inversely proportional to the quality of
information stocks (Tribelsky and Sacks, 2011). For example, when building design proceeds with
outdated information, it can result in incorrect interpretation and ad hoc amendments by the
construction team on-site (Tribelsky and Sacks, 2011; Alencastro et al., 2018).
In the model, inter- and intra-stage information flows are simplified and relate to task work.
Alignment influences information sharing once partners engage in project tasks. Communication
prior to project start is not modelled explicitly. It is assumed that a maximum stock of information
1/"** is required to complete the tasks associated per building area, per stage without defects, and
there is no information overflow effects. It is assumed that intra-stage communication flow C;
increases with alignment Ai, and the rate of aggregate partner engagement Ei per month. C; is given
by:
C; = min(E; x Aj, i" — 1,) (3)
1; relates to the amount of change in partner understanding, which is extremely difficult to identify
and measure (Daft and Macintosh, 1981). /; is thus defined as the quantity of data that is gathered
and interpreted by organization participants i.e. it represents an information stock. Project partners
make sense of a project and work towards its delivery as they cope with ambiguity, uncertainty and
complexity (Weick, 1995). Inevitably some quantitative information will tend to become out of date
as the project progresses i.e. information has a half-life (Samset and V olden, 2016). It is assumed
that intra-stage information Ij accumulates with C;, and erodes inversely proportional to Ai, and
stage duration D;. Ii is given by:
h=Jy(G-za5)at (4)
The reciprocal nature of information exchange between stages i and j suggests a multiplicative
relation. It is assumed that inter-stage communication C;; increases with A,;, C;, and C; and is given
by:
Cy = min(C; x G x Ai, LP — ly ) (5)
The stock of inter-stage information J;; is assumed to erode when project stage ends, and project
details are stored away. J;; depends on C;;, stage specific erosion depends on duration D; and is
given by:
t Cy
ly = Sy (cy - at) dt (6)
4.2. Project Work, Control and Rework
The resources K; in stage i are dynamic and depend on its duration D; and the total net tasks to be
completed T;. K; is assumed to represent full time employees, that are reallocated to other projects
when 7; declines and falls below K;. It is given by:
K, = fo(T/Di — (K, - T)) at (7)
Project work rate on tasks per building area a, encompasses task completion, quality assurance, and
rework, and is subject to K;. Suppressing a for clarity, the task completion rate R; per building area
for stage i is given by:
Ri = min(T, Ki)/ti (8)
Where t;, is the time required for task completion. Quality assurance Q; is influenced by the
alignment A; of state i partners to high quality work, and is given by:
Qi = max(0, Kj — Ri) X Ai (9)
Rework in projects is work that has to be repeated and can arise from defects in project execution or
from client requirement changes (Love and Edwards, 2004). Defects may arise in any stage, from
unrealistic design programmes, organizational culture, quality assurance practices, changes of client
needs, and a lack of a common language with which to articulate client requirements in design stage
that could lead to misalignment and unnecessary amendments by teams working on-site in
subsequent stages (Lopez et al., 2010; De Wilde, 2014; Alencastro et al., 2018). Defects range from
few to several hundred, and several kinds of defect classification exist (Alencastro et al., 2018).
Tasks are also assumed to be small enough to be defective or correct but not partially defective
(Ford and Sterman, 1998)°. The quantity of tasks rework in design and construction stages is
inversely proportional to the quality of information stocks, which is assumed to increase with
quantity /;; (Tribelsky and Sacks, 2011). The rate of defect generation G; in stage i depends on the
rate of task completion R;, the stage contribution P; to defects that affect building quality. It is
assumed that inter-stage information exchange /;; provides the necessary detail to complete tasks
and reduce G; per building area in stage I normalized by the total number of tasks per building area,
or scope T;orq: Which is assumed to be the same for every building area in every stage. G; is given
by:
Gi = Ri XP, X (1 — lij/Trotar) (7)
The discovery rate of intra-stage defects F; in stage i depends on quality assurance test Q; and is
subject to resource constraints. F; depends also on the number of completed tasks to test T;;, the
level of defect testing thoroughness H;, and the contribution of stage i to defects P;, The defect
discovery rate F; is given by:
F, = min(Q;, Tri X Hi X Pi) (8)
Defects in one stage are detected often in later stages, where they have some knock-on effect,
(Sommerville, 2007; Aljassmi and Han, 2013; Alencastro et al., 2018). For example, defects arise
frequently in the design stage with mis-communication between client and design team, or between
the members of the design team, about building performance targets (De Wilde, 2014). These
defects may be discovered by the main contractor in the construction stage through quality
assurance. The defects that are discovered in stage j and attributed to defects in previous stage i
depend on the proportion of defects to tasks P,; that flow from stage i to j, and the proportion k, of
defects possible to rework in stage j. Fj, is given by:
Fy = min(Tpj,Q; — Fj x Pij X Hj) x (1— ky) (9)
It is assumed that intra- and inter-stage information (eq. 4, 6) can increase quality testing
thoroughness H,, and the probability of defect discovery given the initial scope T;o¢q, (Tribelsky
and Sacks, 2011).
H; = min (1, (Ho, + (i x hij)/Teeat)) (10)
Where Ho; is the initial probability of defect discovery. Nevertheless, some known defects in each
stage may not be corrected as most partner resources are reassigned to other projects due to resource
and time shortages during a project stage (Love et al., 2002). This reduces project partner capacity
to receive tasks for rework, and makes more likely the use of workarounds. This resource shortage
effect follows an s-curve® and is modelled with a standard logistic s-curve S; for each stage j with
value range (0..1) (Sterman, 2000). 5; accounts for resource, costs, time pressure related effects that
are not modelled explicitly due to insufficient information, but is also used to simplify the model.
8 This assumption also becomes more accurate as task size becomes smaller.
° Macleamy, P. (2004). Collaboration, Integrated Information, and the Project Lifecycle in Building Design,
Construction and Operation. The Construction Users RoundTable.
http://www lcis.com.tw/paper_store/paper_store/CurtCollaboration-20154614516312.pdf (accessed 16/1/2018)
The rate of intra stage defect correction is based on Ford and Sterman (1998)?” is multiplied by
(1 —5;) to account for resource related stage constrains. The inter-stage retum rate of defective
tasks Rj; from stage j to i depends on k,, the inter-stage alignment A; ; and S;. Rj; is given by:
Ry = Ai X Tri X 1 - Si)/ti (12)
Where ¢;; is the return delay from stage j to i. As S; becomes 1 all remaining known defects flow
downstream to account for knock on effects on final building quality. The final quality of a building
area relative to design targets is assumed to be directly proportional to the ratio of defects over the
number of project tasks related to the building area. This ratio provides the quality deviation of
building areas from their baseline design operational quality, and is the basis for the interface with
the building performance model.
4.3 Interface with the Building Performance Model
The SD project management model interfaces with the building performance model developed in
Design Builder simulation software with Energy Plus© as the simulation engine!!. The SD model
produces a quality index output for the building areas with known performance issues (Table 1).
This facilitates the interface between the SD and building performance model and the expert
knowledge elicitation process about the H; and P; variables.
An example of a building area with lower performance than the design target is the heating
system efficiency. It is low due to the under sized heating terminals and malfunctioning heat pumps.
One element of the task and defect arrays in the SD model is used to trace heating system quality
through the project stages. The final quality deviation for the heating system is used as the input in
the Design builder model. The Design Builder input parameter is the heating system Coefficient of
Performance (COP). It represents the aggregate effect of the heating system issues on the building
performance. Table 1 shows the correspondence of SD array elements and Design Builder input.
Table 1 Identified performance issues related to project defects in the case building
sD as o142
Array Building Energy Plus Actual Building Remarks
Area Input Defect
Element
. Undersized heating
Heating COP value of terminals, issues with COP represents the
1 System heati ; aggregated system
ae eating system heat pumps in hot
Efficiency performance
water vessels
Lighting Lighting Load _ Increased lighting '
2 power density per unit area load than designed Direct Input
Office Office Increased small
3 equipment Equipment Load power load than Direct Input
power density per unit area designed
Number of
4 Occupant People per unit Increased number of Direct Input
density people than designed
area
10 See eq. 30 in the working paper version of Ford and Sterman (1998), available from:
https://dspace.mit.edu/bitstream/handle/1721.1/2644/SW P-3943-36987273.pdf?sequence=1 (accessed16/01/2018)
11 A simplified, validated version of the building performance model is used to reduce computation time (approximately
24 hours for a single run)
10
Building operating at Direct Input (set point
Heating Set — Heating system, higher Temperatures maintained during
pout Set point than designed occupied hours)
Oxeananie Oveipane Building used for Direct Input (hours of
6 Panty peney longer hours than weekday occupancy
hours Schedule hours :
designed changed).
Manually operated
7 Infiltration Infiltration Rate vents not shut Direct Input
always/properly
Ventilation Faulty sensors inthe Sensor defects can be
Control: CO, building leading to —_represented by changes
: increased CO2 in CO: concentration
concentration control.
8 Ventilation
Concentration
The SD model development for the case building benefitted and used input from two industry
experts involved in the research. Both had access, and discussed with multiple project partners
(architects, engineers, contractors), they visited the building and conducted four rounds of
interviews with the facilities management team in 2016-2017. The experts also developed and
calibrated the building physics model of the case building and thus had in-depth knowledge of the
performance gap areas in the building. The experts through the analysis they conducted in
EnergyPlus ©, pointed to building areas with performance issues and this was used to set number of
task and defect arrays in the SD model. Based on their knowledge of the case and their prior
experience, they provided expert judgement and input estimates for the SD model on the testing
thoroughness of building development quality in each stage, and the contribution of each project
stage to the building areas where a performance gap had been identified (A ppendix B).
5 Model Simulation
5.1 Model calibration and testing
The SD model uses the following inputs: (i) expert estimate range on the contribution of each
project stage P; to the end quality of the building, (ii) expert estimate range on quality assurance
thoroughness H; at each stage, (iii) work concurrency in, and between stages", (iv) difficulty in
making task related changes S; in each project stage, (v) the proportion of upstream defects that are
reworkable in downstream stages k;, (vi) performance gap figures established for the case building
through building performance modelling and analysis, and (vii) level of initial alignment. A ppendix
B provides tables for (i) and (ii), (iii) is given a value of 90% based on expert judgement so that
90% of stage related work has to be completed for downstream stage work to begin, (iv) is set
through expert judgement (see A ppendix C for details), (v) was set to a value of 0 based on expert
judgement on the project, (vi) was provided by building performance analysis (Jain et al., 2017),
and (vii) was elicited through project partner interviews.
Model calibration was carried out through numerical optimization to estimate model
parameters that minimize SD model output error to performance gap data (Oliva, 2003). The SD
model quality index output for building areas with an identified performance gap was less than data
based on the building physics model which is calibrated on building monitoring data. It is assumed
that knock-on effects have a greater than unit effect, in line with theory (Lyneis and Ford, 2007) and
! For concurrence relations see fig. 6-8 in Ford and Sterman (1998).
11
evidence (Parvan et al., 2015). The knock-on effect of defects N,; from stage i to j per building
performance area depends on the sum of undiscovered defects T,,; and known defects 7;
normalized against initial design scope T;o¢q:, and the strength y of knock on effect. Values for y are
obtained by minimizing simultaneously the performance gap error of model output for each
building area (see A ppendix C for details). Nj; is given by:
Ni = (1+ Cui + Tri)/T rota)” * (12)
5.2 Simulation Results
To characterize the range of behavior the system produces and to understand the impact of each of
its parameters, the model has heen extensively analysed. A range of plausible scenarios has been
explored to highlight the management and operational trade-offs in such projects. The two
independent building performance experts provided input to the model in the form of minimum,
maximum, and best estimates for the contribution of each stage to end building quality P;, and
testing thoroughness H; in each stage. The input space of their best estimates was explored in 729
runs, to produce the output of the possible, average building quality of the nine building areas
relative to the maximum value of design building performance of one (Figure 5). The number of
tasks for each building area T;¢q, is set to 100. Simulation time is five years (see project timeline
in Appendix D). All average quality curves illustrate the accumulation of defects that reduce
building quality and cause the initial, narrow building quality range to widen. Quality rises as
defects are reworked in each stage. Figure 5 broadly reveals that most quality gains or losses are
made in construction stage.
Building Quality
2000 2001 2002 2003 2004 2005
Year
Figure 5 Building quality results using expert best estimates!?
The breakdown of building performance deviation from design targets in the areas where a
performance gap has been observed in reality is shown in Figure 6"*, On all the defect categories the
range of expert input used for produces a min-max range (shaded grey bars) that envelopes the
actual real performance (black line). The average value of expert estimates underestimates quality
on some building area and overestimates it in others thus there is no clear evidence of bias error.
13 The actual years of construction have been changed to preserve anonymity.
\ The ninth area is omitted as performance has been restored to design targets by the end of the 2™ year post
commission.
12
zzz Min
mmm Real Estimate
Calibrated Model
mmm Expert 1
imme Expert 2
zzz Max
Energy Performance Deviation
”
Building Performance Areas
Figure 6 Performance simulation results for building areas with underperformance
The SD output in Figure 6 is the input to the building performance model to generate its
corresponding annual energy performance. Figure 7 shows the total energy consumption including
electricity and gas (kWh/m’). Expert 1 results are below min because his input concerns 7 building
performance areas (see A ppendix B)'°, so the total building performance is better.
Expert 1
—\ Z Best
Estimate
Expert 2
‘i Best
SE oer Estimate
oes &§ Po ees Ee $
$ § € FFE FY FS é #
§ ¥ ° g é Zé
Total Energy Consumption (kWhim?)
o¢
Figure 7 Total, monthly energy consumption from building performance model
5.3 The Effect of Alignment on Performance
Tests for increased, initial CSC alignment in line with UK reports (Latham, 1994; Egan, 1998)
show its effect on average building quality (Figure 8, left). Raising intra and inter stage initial
alignment from 0 to 2 does result in modest improvement of average building quality 17.67%. The
effect of increased inter-stage alignment only is hampered by the level of intra stage alignment that
affects intra-stage work quality. However, with resource constraints implemented in both cases, the
variation in alignment does not translate into significant building performance figures. The
improvement in average building quality is higher when S-curve resource constraints across the
three stages are removed and alignment is raised from 0 to 2 (31.05%).
An alignment value of 1 is assumed to be the maximum that a CSC can operate under. A
value of 1 in initial alignment with no resource constraints reduces total annual electricity cost by
2.2% and CO2 emissions by 2.6% compared to the zero-alignment case (Table 2). However, the
removal of resource constraints leads also to 4.48% increase in total tasks reworked across all
15 Expert 1 did not provide estimate for infiltration because he did not have access to the calibrated model for the case
building.
13
stages. This rework needs additional resources and cost, if the project is to be delivered in time.
Most of the additional work concerns construction stage tasks and it is done when the project is
already in the operation stage. This insight provides some supporting evidence for soft landings
approach implemented currently in the UK (De Wilde, 2014). The building performance results
suggest that project managers should attend to alignment and resources during project planning
since they are critical to success to the project, but they generate also energy use savings and CO2
emission reductions with more work.
Total Energy Consumption (kWhim?)
< oe :
SEPP HES SS SF SS
— s-Cure
No §-Curve Calibrated Run
ter-Stage Alignment with S-Curve ——— Initial Alignment=0
Inter Stage Alignment without S-Curve Initial Alignment=1
Figure 8 The effect of initial alignment (left) on total energy consumption (right).
Table 2 Annual building performance in scenarios with initial alignment of 1
Annual Calibrated Initial Initial
Performance mun alignment=0 alignment=1
Energy cost (£) 54005.04 57836.18 56555.71
CO2 emissions (kg) _223808.2 237942.1 231768.3
Work done (tasks) 3066.36 3074.33 3212.07
Figure 8 shows the reinforcing effect of alignment and resource availability, and raises the issue of
how to achieve this in practice through appropriately designed contracts that incentivize CSC
partners. These results echo the trade-off of the sustainability team leader, who estimated that the
time and human resources required for a complete energy modelling study was just not enough,
although it was the consensus opinion that it would make a difference in quality. The corresponding
results from the building performance model show the performance gains when initial alignment
between partners is 1 (Figure 8, right). It results in energy consumption savings and CO2 emission
reduction (Table 2).
5.4 Early Engagement Scenario
Building energy performance is an outcome that arises from complex interactions of building
elements and occupant behavior. The inclusion of a high energy performance goal in the building
case represents an increase in project complexity relative to the norm for buildings of this type. This
raises the need to develop a shared understanding of project targets and communication between
project partners (Hong et al., 2004). Clear project targets that are well communicated, understood
and accepted improve overall teamwork because project team members engage in goal related
functions.
14
The soft landings approach followed in the case building is designed to keep designers and
constructors involved in the performance of buildings beyond completion (De Wilde, 2014). Project
partners stay engaged and exchange information, while physical work in each stage may end. A key
point for successful team and project development performance is the project team building process
around a set of targets such as quality, cost and development time (Hong et al., 2004). An early
engagement scenario is simulated where project partners initiate interactions that alignment and
communication. Early alignment and information sharing between project partners facilitates clarity
on project targets and enables performance.
This scenario was tested by varying information sharing and early engagement between
partners in the model. The reference case runs use the actual project timing from the case where
partners in design, construction and operation stages engage and start work in month 0, 11.5 and 33
(see Appendix D). It is assumed that project engagement and information sharing is distinct from
physical work and can thus start earlier in the project. In this scenario partner engagement begins in
month 0 while physical project stage work begins in months 0, 11.5, 33.
The results show that earlier engagement and communication is beneficial to building quality and
building performance in building areas (Figure 9). This in support of prior research on the extent of
front-end development activities and their influence on project performance where lack of maturity
in project definition prior to project execution proved to be responsible for the failure of major
projects (Suprapto et al., 2015). The corresponding output from the building simulation model
shows a significant reduction in average annual energy consumption of maximum reduction in costs
is 28.37% in emissions is 29.25%.
25 "1
< 10
§ z=
220 Zo
3 =
& S68
gis e
; a
S10 48
é ?
S ge
205 a
fi Ba
2
00 L °
; £ Ae s $ so §
: a ee
aulding Pert cal oF
mmm Calibrated Run, Alignment = 0.5 — Callorates Run, Alignment = 0.5
jam Early Engagement, Alignment = 0.5 —— Early Engagment, Alignment = 0. is
Early Engagement, Alignment = 1 Early Engagement, Alignment =
Early Engagement, Alignment = “no resource constraints Ear Empat Agimont =n escue constants
Figure 9 Effect of partner engagement and communication on average building quality (left), and
energy performance of building areas.
Table 3 Annual building performance in engagement scenarios with initial alignment of 1
Early engagement,
Annual Calibrated Early Early Alignment = 1, no
Performance mn engagement, engagement, resource
Alignment =0.5 Alignment =1 eonsiraints
Energy cost (£) __54005.04 50959.10 40797.07 38682.01
CO2 emissions 211053.45
(kg) 223808.2 168625.7 158348.2
Work done 3000.74
(tasks) 3066.36 2940.62 3139.76
15
6 Discussion and C onclusions
6.1 Theoretical and Methodological C ontribution
The multi-methodology framework developed in this paper contributes to the system dynamics
literature on project management. The theoretical contribution of the paper is the integration and
operationalization of project partner alignment and information flows in the standard, multi stage
project management model (Ford and Sterman, 1998). This facilitates the exploration of
collaboration related effects on CSCs that are hampered by fragmentation in the UK and elsewhere.
The model structure can facilitate the assessment of project collaboration related effects on building
operational performance and CO: emissions, a quite topical issue in lieu of climate change, that has
been neglected in system dynamics literature. The documentation of the framework and its
illustrative application provides a basis to tackle a class of problems rather than a single case
(Forrester, 1961). The modelling framework is a first step to explore such effects in more cases to
produce generalizable results.
The methodological contribution that enables detailed assessment of operational performance
is the novel integration of SD and building physics methodologies that couples the SD project
management model to a building physics model. In doing so, the multi-methodology framework
seeks simultaneously theoretical generality and situational grounding, while being methodological
rigorous and practically relevance to both fields (Ketokivi and Choi, 2014). Data availability on
project time and cost will permit a replication of the method in the UK context and the building
physics modelling will enable an assessment of the long-term effects of building projects. The
intended aim is to produce research that will alter the way industry insiders look at CSC
collaboration so that CSC partners then consider seriously mechanisms that permit sufficient, timely
and accurate information for CSC governance and building operational performance.
6.2. Practical contributions
The motivation for this paper was the share of the UK building sector to total energy consumption
and CO: emissions, and its contribution towards the 80% emission reduction target set by the UK
government for 2050. This required an explicit focus on the project process that delivers buildings
and an in-depth analysis of the implications for operational building performance. Our approach is
of particular benefit in building project contracts that include energy performance targets. This is a
trend that is picking up pace in the UK and globally (Sorrell, 2007).
The managerial implications of the simulation results are in line with insights from prior work
on project performance in terms of cost, time and quality. An early focus on project performance
and energy targets is important in terms of operational building energy performance. Encouraging
early problem discovery and instituting root cause analysis, energy testing, and capabilities that
facilitate quality, can help set the right level of alignment and effectiveness across the CSC. As the
set of project uncertainties involved expands to include energy specific targets the project must
achieve, project partners and managers should be flexible in updating initial plans when required,
and delegate more responsibility to those on the frontlines who often have a more nuanced
understanding of the actual tasks, performance and quality.
The challenge in adopting such behaviour in actual operations is that conventional project
management performance metrics and tools ignore some of the soft variables and feedback
mechanisms that are explored in this model (Browning, 2010). Moreover, the worse-before- better
trade-offs involved in upfront quality, organizational capability investments, and cost make it harder
16
to learn and pursue the more flexible learning focused style of project management (Repenning and
Sterman, 2002; Williams, 2008). Moreover, building performance improvements cannot depend
only on the voluntary learning of organizations in the building industry. Projects are complex
entities and leaning from complex systems needs a more sophisticated approach than simply
writing down lessons (Williams, 2003). Organizations might have procedures for learning lessons
from projects but few might adhere to those if they don’t perceive immediate benefits in the market,
and the transfer of lessons within an organization is one of the major difficulties of learning from
projects (Williams, 2008). There is lack of time, management support, and incentive to do so.
6.3 Limitations and Future Work
The study has some data and methodological limitations, and some potential for future development
work. The case building was commissioned before it became part of the research project. This
limited access to some project partners e.g. the building architect was not interviewed due to new
project commitments. It also limited data availability with respect to a number of areas: total project
task figures, tasks per building performance area, total resources per stage (due its multi
organizational nature), and partner resource prioritization and allocation vis a vis other projects.
Expert judgement was used to calibrate s-curves for each stage and account partially for these
limitations. The real building performance is known in detail through in-situ monitoring and
building performance modelling, so it is possible to claim that resource related quality effects in
each stage have been captured, albeit implicitly in expert estimates on quality and testing
thoroughness. Accurate resource availability information would increase the realism of the
retrospective analysis and enable a better assessment of the information exchange and collaboration
effect on building quality.
One way to overcome such limitations in future framework applications is to follow a
building project from its inception to its completion through a process tracing research design
(Collier, 2011; Bennett and Checkel, 2014). This would forego the reliance on interviews and post
hoc estimates, but most importantly it would increase managerial relevance as it would take on
board manager’s perspectives explicitly at the outset of research project (Holmstrom et al., 2009).
Acknowledgements
The authors gratefully acknowledge the financial support from ‘The 'Total Performance! of Low
Carbon Buildings in China and the UK’ (‘TOP’) project funded by the UK EPSRC (Grant code:
EP/N009703/1) Corresponding research carried out in China is funded by NSFC China
(51561135001).
References
Adner, R., 2017. Ecosystem as structure: An actionable construct for strategy. Journal of
Management 43, 39-58.
Alencastro, J., Fuertes, A., de Wilde, P., 2018. The relationship between quality defects and the
thermal performance of buildings. Renewable and Sustainable Energy Reviews 81, 883-894.
Aljassmi, H., Han, S., 2013. Analysis of causes of construction defects using fault trees and risk
importance measures. Journal of Construction Engineering and Management 139, 870-880.
Aljassmi, H., Han, S., Davis, S., 2016. Analysis of the complex mechanisms of defect generation in
construction projects. Journal of Construction Engineering and Management 142, 04015063.
Baiden, B.K., Price, A.D.F., 2011. The effect of integration on project delivery team effectiveness.
International Journal of Project Management 29, 129-136.
17
Barratt, M., 2004. Understanding the meaning of collaboration in the supply chain. International
Journal of Supply Chain Management 9, 30-42.
Bendoly, E., 2014. System dynamics understanding in projects: information sharing, psychological
safety, and performance effects. Production and Operations Management 23, 1352-1369.
Bendoly, E., Swink, M., 2007. Moderating effects of information access on project management
behavior, performance and perceptions. Journal of Operations Management 25, 604-622.
Bennett, A., Checkel, J.T., 2014. Process Tracing. Cambridge University Press, Cambridge,
England.
Briscoe, G.H., Dainty, A.RJ., Milet, S.J., Neal, R.H., 2004. Client led strategies for construction
supply chain improvement. Construction Management and Economics 22, 193-201.
Browning, T.R., 2010. On the alignment of the purposes and views of process models in project
management. Journal of Operations Management 28, 316-332.
Cohen, R., Bordass, B., 2015. Mandating transparency about building energy performance use.
Building Research & Information 43, 534-552.
Cohen, R., Standeven, M., Bordass, B., Leaman, A., 2001. Assessing building performancein use 1:
the Probe process. Building Research & Information 29, 85-102.
Cohen, S.G., Bailey, D.E., 1997. What makes teams work: Group effectiveness research from the
shopt floor to the executive suite. Journal of Management 23, 239-290.
Collier, D., 2011. Understanding process tracing. Policital Science and Politics 44, 823-830.
Daft, R.L., Macintosh, N.B., 1981. A tentative exploration into the amount and equivocality of
information processing in organizational work units. A dministrative Science Quarterly 26, 207-224.
Davies, M., Oreszczyn, T., 2012. The unintended consequences of decarbonising the built
environment: A UK case study. Energy and Buildings 46, 80-85.
DBEIS, 2017. 2016 UK Greenhouse Gas Emissions, provisional figures. Department for Business,
Energy, and Industrial Strategy London, UK.
De Wilde, P., 2014. The gap between predicted and measured energy performance of buildings: A
frameworkd for investigation. Automation in Construction 41, 40-49.
Dietrich, P., Eskerod, P., Dalcher, D., Sandhawalia, B., 2010. The dynamics of collaboration in
multipartner projects. Project Management Journal 41, 59-78.
Egan, J., 1998. Rethinking Construction. Department of the Environment, London.
EIA, 2017. International Energy Outlook 2017. U.S. Energy Information Administration.
Ford, D.N., Sterman, J.D., 1998. Dynamic modeling of product development processes. System
Dynamics Review 14, 31-68.
Forrester, J.W., 1961. Industrial Dynamics The M.1.T. Press, Cambridge, Massachusetts.
Frohlich, M.T., Westbrook, R., 2002. Arcs of integration: an international study of supply chain
strategies. Journal of Operations Management 19, 185-200.
Galbraith, J.R., 1973. Designing Complex Organizations. Addison-Wesley Longman Reading, MA.
Gattoma, J.L., 2009. Dynamic Supply Chain Alignment: A New Business Model for Peak
Performance in Enterprise Supply Chains Across All Geographies. Gower Publishing, Farnham,
UK.
Gram-Hanssen, K., Georg, S., 2018. Energy performance gaps: promises, people, practices.
Building Research & Information 46, 1-9.
Han, S., Love, P., Pefia-Mora, F., 2013. A system dynamics model for assessing the impacts of
design errors in construction projects. Mathematical and Computer Modelling 57.
Hanson, J.D., Melnyk, S.A., Calantone, R.A., 2011. Defining and measuring alignment in
performance management. International Journal of Operations and Production Management 31,
1089-1114.
Hedstrém, P., Swedberg, R., 1998. Social Mechanisms: An Analytical Approach to Social Theory.
Cambridge University Press.
Hensen, J., Lamberts, R. (Eds.), 2011. Building Performance Simulation for Design and Operation.
Spoon Press, London.
18
Hoegl, M., Gemuenden, H.G., 2001. Teamwork quality and the success of innovative projects: a
theoretical concept and empirical evidence. Organization Science 12, 435-449.
Holmstrom, J., Ketokivi, M., A-P., H., 2009. Bridging practice and theory: A design science
approach. Decision Sciences 40, 65-87.
Holtz, G., Alkemade, F., de Haan, F., Kéhler, J., Trutnevyte, E., Luthe, T., Halbe, J., Papachristos,
G., Chappin, E., Kwakkel, J., Ruutu, S., 2015. Prospects of modelling societal transitions: Position
paper of an emerging community. Environmental Innovation and Societal Transitions 17, 41-58.
Hong, P., Nahm, A.Y., Doll, WJ., 2004. The role of project target clarity in an uncertain project
environment. International Journal of Operations and Production Management 24, 1269-1291.
Howick, S., Ackerman, F., 2011. Mixing OR methods in practice: Past, present and future
directions. European Journal of Operational Research 215, 503-511.
IPCC, 2014. Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group
III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change In:
Edenhofer, O., Pichs-Madruga, R., Sokona, Y ., Farahani, E., Kadner, S. , Seyboth, K., Adler, A.,
Baum, I., Brunner, S., Eickemeier, P., Kriemann, B., Savolainen, J., Schlémer, S., von Stechow,
C., Zwickel, T., Minx, J.C. (Ed.). Cambridge University Press, Cambridge, United Kingdom and
New Y ork, NY, USA.
Jain, N., Burman, E., Mumovic, D., Davies, M., Tindale, A., 2017. Improving the energy
performance contracting process by managing technical risks and uncertainties using building
performance simulation: Lessons from a case study in the UK. The 15th International Conference of
International Building Performance Simulation Association, San Fransisco, USA.
Jingmond, M., Agren, R., 2015. Unravelling causes of defects in construction. Construction
Innovation 15, 198-218.
Kache, F., Seuring, S., 2014. Linking collaboration and integration to risk and performance in
supply chains via a review of literature reviews. International Journal of Supply Chain Management
19, 664-682.
Ketokivi, M., Choi, T., 2014. Renaissance of case research as a scientific method. Journal of
Operations Management 32, 232-240.
Kohler, J., De Haan, F., Holtz, G., Kubeczko, K., Moallemi, E., Papachristos, G., Chappin, E.,
2018. Modelling sustainability transitions: An assessment of approaches and challenges. Journal of
Artificial Societies and Social Simulation 21.
Latham, M., 1994. Constructing the Team: Joint review of Procurement and Contractual
Arrangements in the United Kingdom Construction Industry. HMSO, London.
Lee, H.L., Padmanabhan, V., Whang, S., 1997. Information distorion in a supply chain: The
bullwhip effect. Management Science 43, 546-558.
Lopez, R., Love, P.E.D., Edwards, DJ., Davis, P.R., 2010. Design error classification, causation,
and prevention in construction engineering. Journal of Performance of Constructed Facilities 24,
399-408.
Love, P.E.D., Edwards, D.J., 2004. Forensic project management: The underlying causes of rework
in construction projects. Civil Engineering and Environmental Systems 21, 207-228.
Love, P.E.D., Edwards, D J., Irani, Z., 2008. Forensic project management: An exploratory
examination of the causal behaviour of design-induced errors. IEEE Transactions in Engineering
Management 55, 234-247.
Love, P.E.D., Holt, G.D., Shen, L.Y., Li, H., Irani, Z., 2002. Using system dymamics to better
understand change and rework in construction project management systems. International Journal of
Project Management 20, 425-436.
Love, P.E.D., Irani, Z., Edwards, D.J., 2004. A seamless supply chain management model for
construction. International Journal of Supply Chain Management 9, 43-56.
Lowe, R.J., 2007. Technical options and strategies for decarbonizing UK housing. Building
Research & Information 35, 412-425.
Lyneis, J.M., Ford, D.N., 2007. System dynamics applied to project management: a survey,
assessment, and directions for future research. System Dynamics Review 23.
19
Meng, X., 2013. Change in UK construction: Moving toward supply chain collaboration. Journal of
Civil Engineering and Management 19, 422-432.
Mingers, J., Brocklesby, J., 1997. Multimethodology: towards a framework for mixing
methodologies. Omega 25, 489-509.
Mingers, J., White, L., 2010. A review of the recent contribution of systems thinking to operational
research and management science. European Journal of Operational Research 207, 1147-1161.
Molenaar, K., Songer, A.D., Barash, M., 1999. Public-sector design/build evolution and
performance. Journal of Management Engineering 15, 54-62.
Morrison, B., 2015. The problem with workrounds is that they work: The persistence of resource
shortages. Journal of Operations Management 39-40, 79-91.
O’Brien, W.J., Formoso, C.T., London, K.A., Vrijhoef, R., 2009. Introduction. In: O’Brien, W.J.,
Fromoso, C.T., Vrijhoef, R., Kerry, L.A. (Eds.), Construction supply chain management handbook.
Taylor & Francis Group, London.
Oliva, R., 2003. Model calibration as a testing strategy for system dynamics models. European
Joumal of Operational Research 151, 552-568.
Oreszczyn, T., Lowe, R., 2010. Challenges for energy and buildings research: objectives, methods
and funding mechanisms. Building Research & Information 38, 107-122.
Papachristos, G., 2012. Case study and system dynamics research: Complementarities, pluralism
and evolutionary theory development. International Conference of the System Dynamics Society, St
Gallen, Switzerland
Papachristos, G., 2014a. Towards multi-system sociotechnical transitions: why simulate.
Technology Analysis and Strategic Management 26, 1037-1055.
Papachristos, G., 2014b. Transition inertia due to competition in supply chains with
remanufacturing and recycling: A system dynamics model. Environmental Innovation and Societal
Transitions 12, 47-65.
Papachristos, G., 2017. Diversity in technology competition. The link between platforms and
sociotechnical transitions. Renewable and Sustainable Energy Reviews 73, 291-306.
Papachristos, G., 2018. A mechanism based transition research methodology: Bridging analytical
approaches. Futures 98, 57-71.
Papachristos, G., Adamides, E., 2016. A retroductive systems-based methodology for socio-
technical transitions research. Technology Forecasting and Social Change 108, 1-14.
Parvan, K., Rahmandad, H., Haghani, A., 2015. Inter-phase feedbacks in construction projects.
Journal of Operations Management 39-40.
Powell, T.C., 1992. Organizational alignment as competitive advantage. Strategic Management
Journal 13, 119-134.
Rahmandad, H., Hu, K., 2010. Modeling the rework cycle: capturing multiple defects per task.
System Dynamics Review 26, 291-315.
Repenning, N.P., Sterman, J.D., 2002. Capability traps and self confirming attribution errors in the
dynamics of process improvement. Administrative Science Quarterly 47, 265-295.
Samset, K., Volden, G.H., 2016. Front-end definition of projects: Ten paradoxes and some
reflections regarding project management and project governance. Intemational Journal of Project
Management 34, 297-313.
Shrubsole, C., Hamilton, I.G., Zimmerman, N., Papachristos, G., Broyd, T., Burman, E., Mumovic,
D., Zhu, Y., Lin, B., Davies, M., 2018. Bridging the gap: the need for a system thinking approach in
understanding and addressing energy and environmental performance in buildings. Indoor and Built
Environment.
Shrubsole, C., Macmillan, A., Davies, M., May, N., 2014. 100 Unintended consequences of policies
to improve the energy efficiency of the UK housing stock. Indoor and Built Environment 23, 340-
352.
Sommerville, J., 2007. Defects and rework in new build: an analysis of the phenomenon and
drivers. Structural Survey, 391-407.
20
Soosay, C.A., Hyland, P.W., Ferrer, M., 2008. Supply chain collaboration: capabilities for
continuous innovation. International Journal of Supply Chain Management 13, 160-169.
Sorrell, S., 2007. The economics of energy service contracts. Energy Policy 35, 507-521.
Sterman, J.D., 2000. Business dynamics: Systems thinking and modelling for a complex world.
McGraw-Hill, New Y ork, US.
Suprapto, M., Bakker, H.L.M., Mooi, H.G., 2015. Relational factors in owner-contractor
collaboration: The mediating role of teamworking. International Journal of Project Management 33,
1347-1363.
Tribelsky, E., Sacks, R., 2010. Measuring information flow in the detaield design of construction
projects. Research in Engineering Design 21, 189-206.
Tribelsky, E., Sacks, R., 2011. An empirical study of information flows in multidisciplinary civil
engineering design teams using lean measures. Architectural Engineering and Design Management
7, 85-101.
Turner, J.R., Miller, R., 2003. On the nature of the project as a temporary organization.
International Journal of Project Management 21, 1-8.
Tushman, M.L., Nadler, D.A., 1978. Information processing as an integrating concept in
organizational design. Academy of Management Review 3, 613-624.
Vrijhoef, R., Koskela, L., 2000. The four roles of supply chain management in construction.
European Journal of Purchasing & Supply Management 6, 169-178.
Weick, K.E., 1995. Sensemaking in organizations. Sage Publishers, London.
Williams, T., 2003. Learning from projects. Journal of the Operational Research Society 54, 443-
451.
Williams, T., 2008. How do organizations leam lessons from projects-and dothey? IEEE
Transactions in Engineering Management 55, 248-266.
Williams, T., Samset, K., 2010. Issues in front-end decision making on projects. Project
Management Joumal 41, 38-49.
Wong, C., Skipworth, H., Godsell, J., Achimugu, N., 2012. Towards a theory of supply chain
alignment enablers: a systematic literature review. Intemational Journal of Supply Chain
Management 17, 419-437.
Yin, R., 2003. Case Study Research: Design and Methods. Sage, Beverly Hills, CA.
Zervos, A., Lins, C., Muth, J., 2010. Rethinking 2050: A 100% renewable energy vision for the
European Union. European Renewable Energy Council.
Zolfagharian, M., Romme, A.G.L., Walrave, B., 2018. Why, when, and how to combine system
dynamics with other methods: Towards an evidence-based framework. Joumal of Simulation.
21
