Design for evolvability at project front-end strategizing
theory and methods
Capital projects involve designing and delivering new assets that are planned to operate for several decades, but over the project and operating lifetimes, design requirements are likely to change. This leads to a crucial problem at the heart of front-end strategizing: how to reconcile budgetary constraints with capital investments in flexibility to ensure that the asset can be adapted economically over time? The analysis of empirical findings in two distinct organisations shows that practitioners use options logic to address this problem, but they do it intuitively. To improve the quality of the conversation in a multi-stakeholder environment and strike an adequate balance between short-term affordability and long-term adaptability, we argue that the process needs to be formalised. We call this design for evolvability. We introduce the notion of the champion of design for evolvability, a representative of the developer in charge of building informal options thinking into design decision-making at front-end strategizing. To support this role, we arm the champion with a proof-of-principle method, which requires project teams to (1) qualify options in light of foreseeable uncertainties; (2) assess alternative design concepts with and without flexibility built-in; and (3), recommend a viable design strategy
Keywords: project front-end strategizing; options logic; design for evolvability; uncertainty; flexibility
A fundamental problem in the world of capital (infrastructure) projects (e.g., projects delivering airports, bridges, or power stations) is the need to design and build assets that can adapt economically to evolution in requirements. These assets may take many years to deliver and are designed to operate for several decades, but during project delivery and service lifetime, the external environment is likely to evolve. External events can be expected to trigger change in functional and operational requirements, which in turn need to be accommodated through design changes. The adaptation costs are a function of the flexibility built in the design architecture at front-end strategizing—the period upfront when the project team conceptualises and compares alternative designs and chooses a concept to progress into the next stage (Miller & Lessard, 2007). Changes in requirements that surface after project completion can be extremely costly and disruptive to implement if the design of the asset has limited flexibility. Ultimately, an asset can become prematurely obsolete if the adaptation cost becomes unaffordable to the organizations that own and use the asset.
The management of the risk that requirements will change over time has been central to the literature in the management of projects (Chapman & Ward, 1998; Lessard & Miller, 2000; Morris, 1994). This literature highlights that front-end strategizing tends to have a disproportionate impact on the quality of the project outcomes in relation to the elapsed time it occupies in the project life cycle (Miller & Lessard, 2007). It also emphasizes that front-end strategizing invariably happens in fuzzy institutional contexts of conflicting interests and priorities among powerful and legitimate project stakeholders (e.g., funders, the developer, often termed the client from a project suppliers’ perspective), operators (or project customers), suppliers, and other relevant stakeholders such as local communities and public agencies. This problem is compounded by difficulties in developing a concept design that reconciles the financial, political, technical, and social drivers shaping the project (Miller & Lessard, 2007). Notwithstanding this, project client teams have still been exhorted to assess and document the likelihood of future changes in requirements and their potential impact on risk registers. These registers inform decisions about contingency planning and investments in risk mitigation (Chapman & Ward, 1998; Miller & Lessard, 2000; Morris, 1994).
The problem of deciding whether to invest in design flexibility at risk, however, is arguably not dealt adequately through conventional risk management practices (Gil, Tommelein, & Schruben, 2006; Gil, Tommelein, Stout, & Garrett, 2005). Design definitions that can cope economically with change often require extra capital investments with long payback periods. These investments may pay off if uncertainties resolve favourably in the future but they will have used up resources that could have been invested elsewhere. Hence, whenever capital is a scarce resource, project teams need to balance upfront investments in flexibility—which can be framed equivalent to buying an insurance policy (Neufville & Scholtes, 2011)—with the risk of incurring costly changes later on if these investments are ruled out upfront. This represents a delicate decision for public agencies whenever they operate on tight budgets and face competing needs to allocate their resources. Private developers face the same problem because they operate under pressure to achieve profits in short periods. Furthermore, this balancing act happens under conditions of uncertainty about future requirements, bounded rationality (Simon, 1962), urgency to make decisions (Miller & Lessard, 2007), and constraints in the envelope of affordability (Gil & Tether, 2011).
Recent developments in practice and theory have started to explore how options logic can contribute to addressing this problem. Options logic posits that strategy can be used to gain advantage under uncertainty. In a transaction, an option gives the buyer the right but not the obligation to take an action in the future (Dixit & Pindyck, 1994). This notion suits decision-making under uncertainty because it enables postponing an investment that would not pay off if uncertainties resolve unfavourably in the future, while leaving open the option to make the investment (‘exercise the option’) if uncertainties resolve favourably. Anecdotal evidence indicates instances in which this logic informed the framing of capital projects. For example, a UK health trust has recently spelled out in the tender documents that the consortiums bidding for designing, building, and operating two new hospitals should factor in their bids the price for designing in pre-specified flexibilities. The costs of building in the options, plus those costs that the trust needs to pay if it wants to exercise them in the future, then become parts of the contract (Lee, 2007). Surprisingly, however, industry studies suggest that capital projects teams seldom receive training on formal options logic (Ford, Lander, & Voyer, 2002; Gil, 2007; Kalligeros, 2006). Research studies focused on how to combine formal options logic, both its lexicon and ways to frame problems, with practices at project front-end strategizing, are also scarce. This gap motivates this study.
We organise this paper as follows. After reviewing theory in real options and managing capital projects, we explain our research method. We then analyse the empirical findings of an exploratory study of a ‘future-proofed'1 viaduct and an embedded case study of project front-end strategizing practice at Network Rail (NR), the owner and operator of Britain's rail infrastructure. We discuss how practitioners almost invariably use options logic to support design decision-making at the project front-end, but do it intuitively. By cross-fertilizing our empirical findings with theory, we propose a proof-of-principle method to support project teams incorporating options logic at front-end strategizing.
BACKGROUND: OPTIONS LOGIC AT PROJECT FRONT-END STRATEGIZING
Project front-end strategizing exhorts developers to invest time and effort at the onset to think through alternative scenarios that might affect design requirements (Morris, 1994). To be effective, project teams need to combine prescriptive activities such as defining the scope and tasks, risk management, and planning contingent actions (Cleland & King, 1983; Cooper & Chapman, 1987) with other activities such as scenario planning, talking to end-users/communities, and discussing the political/economic environment (Lessard & Miller, 2000; Miller & Lessard, 2007; Morris, 1994).
Risk management practice has long been at the core of front-end strategizing (Miller & Lessard, 2001). Broadly, a risk refers to the possibility that a foreseeable event will occur in the future, which may impact the performance of the project process or operations. The nature of the risks is diverse, and may include changes in patterns of demand; emergence of new technologies or issues with novel, not fully tested technologies; evolution in regulatory frameworks, design standards, and codes of practices; and opposition to the project from particular stakeholder groups. One extant typology (Miller & Lessard, 2007) categorises risk as: (1) market-related (due to uncertainty in demand, financial activities, or supply inputs); (2) completion (due to engineering difficulties faced during the design, construction, and operations); and (3) institutional (due to regulatory, social, and governmental issues). Risk management practices aim at identifying risks, assessing their impacts, producing risk registers, and recommending mitigation and contingency plans (Chapman & Ward, 1998). The unifying purpose is to reduce the adverse outcomes (downside risks) and exploit opportunities in favourable scenarios (upside risks). Upfront risk management is not trivial, however, due to uncertainty in future states of the world and sharp heterogeneity across the stakeholders’ interests, knowledge bases, and priorities. Awareness that decisions made at the front-end strategizing can have disproportionate impacts on the quality of the project outcomes has motivated calls for integrating existing risk management practice with options logic (Gil, 2007; Miller & Lessard, 2007).
Options Logic in Decision-making
An option is the right, but not the obligation, to take an action in the future (Dixit & Pindyck, 1994). The use of options logic to frame problems aims at identifying potential investments in flexibility that can help firms benefit from upside scenarios while limiting losses on the downside, i.e., it introduces an asymmetry in the probability of distributions of payoffs (Trigeorgis, 1996). Options logic provides analytical and qualitative methods to assess the investments and help decision-makers make choices to best cope with foreseeable changes in the future.
To date, studies exploring the applicability of options logic to capital projects have predominantly used real options pricing methods to quantitatively assess the value of capital investments with built-in options (Lee, 2007; Wang, 2005; Zhao, Sundararajan, & Tseng, 2004). The quantitative methods draw on theory from financial studies on options, but explore its applicability to problems related to real assets (Amram & Kulatilaka, 1999). All in all, the take-up of this work has been slow in practice for a number of reasons, including: (1) difficulties in making reliable assumptions and ensuring that complex models stay analytically tractable; (2) difficulties in representing real-world problems with those simplified models; and (3) scarcity of skilled resources (Bowman & Moskowitz, 2001; Kalligeros, 2006; Lander & Pinches, 1998).
In the area of technology investments, an alternative research stream—real options reasoning—has gained traction. Rather than seeking precise evaluations of flexible solutions, this research stream draws on options logic to develop qualitative methods and tools aimed at supporting decision-making under uncertainty (McGrath, 1997). McGrath and MacMillan (2000), for instance, present a structured approach that assesses the value of alternative choices based on parameters derived from options logic. These parameters are subdivided into several statements that influence the value of the options, and managers have to score their agreement/disagreement with each statement. Managers’ responses are used to initiate in-depth discussions about the options, uncover potential improvements that would favourably influence option value, and help choose which options to pursue. More recently, MacMillan et al. (2006) combined Discovery Driven Planning and options logic in a method to help firms select and assign resources to competing technology development projects. As each project unfolds, teams are asked to test their assumptions in light of new information and amend previous recommendations. They also need to vary each assumption to investigate the impact on the project value, a metric that compounds the net present value and the option value of the investment.
In the world of R&D projects, Barnett (2005) uses similar ideas to develop a conceptual framework to ask decision-makers to recognize R&D opportunities as options, choose a subset for further development, and ultimately decide on which ones to invest. The framework indicates the right balance between exploration and exploitation in R&D investments. Firms that use the framework diligently can avoid (1) devoting excessive attention to exploitation, which can lead to failure to adjust to alternative market segments and (2) excessively focusing on exploration, which can lead to failure to exploit new R&D opportunities. The use of the framework ensures that managers will intervene in the options realization process, reassessing current and predicted future states, and thus making better-informed decisions in response to changes in market conditions. We next discuss how real options reasoning can also be used to frame problems of design.
Options and Design
Options logic has long been used to frame discussions on product design architectures and flexibility (Sanchez & Mahoney, 1996; Baldwin & Clark, 2000; Brusoni & Prencipe, 2006; Ethiraj & Levinthal, 2004). Product architecture relates to the way through which the functions of a product are allocated to its components (Thomke, 1997; Ulrich, 1995), ranging from modular to integral. Modular architectures build upon Simon's (1962) concept of nearly decomposable systems, where functional elements or subsystems are strongly connected among themselves and weakly connected to other elements or subsystems. Importantly, modularity enables the design to evolve over time because it breaks apart the design interdependencies, allowing each module to be interpreted as an option (Baldwin & Clark, 2000). By mixing and matching modules, designers can alter parts of the system without affecting the whole (i.e., they can exercise their options at limited cost). This flexibility is possible, because each component can succeed or fail independently, and hence the consequences of such alterations become localized.
In some cases, however, integral architectures can be economically more attractive or even the only feasible alternative, as in the case of designing some large-scale infrastructure assets (Gil, 2007). Integral architectures can generate a product with a better overall performance (Fixson & Park, 2008) or that is cheaper to commercialize due to material reductions (Ulrich, 1995); integral architectures can also make it hard for competitors to copy products (Schilling, 2000). Gil's (2007) application of options reasoning to make sense of early decisions in airport terminal design projects shows that options can still be built in integral architectures as long as upfront investments are made in safeguards at risk. Safeguards are investments in design provisions or allowances aimed at enabling to accommodate a foreseeable evolution in requirements at limited cost. They build in a degree of flexibility to cope with change when it is technically too difficult to modularize design architectures, and assets are expected to exhibit long operating lives (e.g., safeguarding the design of underground sewage networks for coping with growth in demand over time). But, like modularization, developers may not be able to afford the upfront investment in safeguarding.
Recent studies have started to explore methods to help project teams assess and compare the value of rigid and flexible designs under uncertainty and decide which one to progress into implementation. Suh, Weck, and Chang (2007), for example, have developed a comprehensive method to support product development teams build flexibility in the design of car platforms—designs expected not to become obsolete for a few decades. This method asks teams to follow seven steps that combine quantitative analysis and expert engineering knowledge: (1) identify uncertainties in functional requirements; (2) relate uncertainty in requirements to specific design elements; (3) identify the set of design elements and their respective required bandwidth that yield maximum revenue; (4) identify the critical design elements that most need flexibility; (5) conceptualise product design alternatives; (6) determine the costs and benefits for each design alternative; and (7) analyse how uncertainty in requirements affects the performance of the design alternatives. Similarly, Kalligeros (2006) proposes a method that integrates Design Structure Matrix (DSM) with real options algorithms to assess investments in flexibility for large-scale systems. The method uses DSM to represent the different functional elements and relationships within a complex system (Eppinger, Whitney, Smith, & Gebala, 1994; Steward, 1981) and explore ways to aggregate strongly interconnected elements into the same modules (Browning, 2001).
Extant literature in management science tends to invariably acknowledge that investments in flexibility can increase the overall product costs. Indeed, neither modularity in product development (Baldwin & Clark, 2000) nor safeguarding in capital assets (Gil, 2007) is free. The exception is Neufiville and Scholtes’ (2011) work, which, interestingly, argues flexible designs can actually be cheaper—and rightly so, because its framing compares the output of the first phase of a capital programme designed in a flexible way with a baseline programme that delivers a rigid alternative size for the long-term operating scenario at risk. Irrespective of the framing, however, extant literature stays short of stimulating a discussion about who ought to pay for upfront investments in flexibility at risk. This is perhaps a lesser issue in the world of product development, where the key negotiations of design decision-making unfold within the boundaries of a single firm. However, the question of who pays for what and why is fundamental in capital project undertakings that require multiple stakeholders to coalesce their own visions at front-end strategizing under conditions of uncertainty and limited capital resources (Gil & Tether, 2011). This is the question at the heart of this study.
Our research method involved an empirical in-depth study on the nature of design decision-making at project front-end strategizing. The fieldwork started with an exploratory study of a £3.5 million award-winning project designed to develop a new viaduct over a river floodplain2. We collected primary data mainly through recorded conversations with the design team at Halcrow Group Ltd. (an engineering consultant) and analysis of archival documents, such as technical reports (Sreeves, 2007), minutes of project meetings, and feasibility studies. We also used investigated clips from the local press and council press releases to learn about the institutional context. Technical notes on other viaduct designs and design standards helped to contrast the characteristics of this particular viaduct with traditional designs. Our aim in this relatively simple project from a stakeholder management perspective was (1) to investigate whether design teams use options logic to inform their decisions when comparing alternative concepts at front-end strategizing, (2) to uncover the level of formality of these decisions, and (3) to assess the complexities and benefits of undertaking a real options pricing analysis.
The insights of the exploratory study motivated us to investigate about the intuitive use of options logic in more complex project environments and how early design decisions unfold in a multi-stakeholder context. To this purpose, we undertook an embedded case study of project front-end strategizing practices at NR. Our unit of analysis was the front-end strategizing processes for three NR capital projects to redevelop existing assets: the £15 million project to provide a new rail chord to connect freight lines in the outskirts of Warrington, the £150 million investment in the revitalization of Reading Station, and the £12 million project to redevelop the overcrowded Salford Crescent Station. Data collection mainly involved semi-structured interviews, analysis of archival documents, and observations at the headquarters of NR Investment Division. Our key informants were programme engineers and environment managers who gave the first author visitor rights to the office and access to the intranet. Between February 2010 and August 2011, we conducted over 30 formal and informal one-on-one meetings with NR staff with diverse job roles (e.g., project manager and engineer, risk manager, commercial sponsor) and with representatives from project clients and other key stakeholders, which we identified through a snowball effect (Biernacki & Waldorf, 1981). We searched the intranet for documents such as project reports, corporate information, capital development procedures, and threads of electronic conversations. We sat in internal meetings (e.g., value management and risk assessment meetings), and we also organised two workshops to receive feedback from preliminary insights from the case studies, research method, and rudiments of the proof-of-principle.
The cross-fertilization of the findings from the empirical fieldwork with literature in the management of capital projects and real options informed the development of a proof-of-principle—the method to design for evolvability—aimed at formalising the use of options logic for early design decision-making. To validate the usability of our method, we transformed the case study on the Salford Crescent Station project into a lab-based simulation exercise. In a follow-up study, we discuss the on-going validation strategy in detail and present some preliminary findings.
‘FUTURE-PROOFING’ CAPITAL DESIGNS AT FRONT-END STRATEGIZING: THE CASE OF THE UPTON-UPON-SEVERN VIADUCT
The Upton-upon-Severn viaduct project aimed at replacing a 1939 reinforced-concrete viaduct over the low-lying flood plain to the east of the River Severn, in Worcestershire, United Kingdom. The 170 metre-long viaduct carries the A4104 highway, and is an important river crossing for local communities. The viaduct was suffering from serious structural problems due to corrosion caused by flood events occurring every five years on average; six complete closures of the road had occurred between 1947 and 2004. The disruptive nature imposed restrictions on weight and traffic and urged the local authority to replace the old viaduct with a new one using the same alignment to avoid statutory delays in planning approval. The project design needed to factor in a life expectancy of 120 years based on published design standards (British Standards, 1988; Highways Agency, 1992). Four alternative concepts surfaced during front-end strategizing (Table 1).
Table 1: Alternative design concepts for Upton-upon-Severn viaduct.
|Alternative||Summary of Assessment of Pros and Cons|
|(1) New viaduct built on site with the deck raised by at least two meters (relative to old one) to safeguard against an extreme 1:100-year design flood event||Risk of delays on getting planning consent to elevate the highway; extra costs for elevating the highway needed to be factored in.|
|(2) New viaduct assembled on site at the same elevation using pre-fabricated concrete culverts||This is a cost-effective and fast concept to execute, but it is aesthetically inferior; the viaduct would need to be replaced if the highway gets elevated in the future|
|(3) New viaduct assembled on site at the same elevation with a multi-span integral design||This is a more costly solution that offers an aesthetically superior design; the viaduct would need to be replaced if the highway gets elevated in the future|
|(4) New viaduct assembled on site with a multi-span modular design and a built-in option to elevate the deck in the future||This solution requires sizing the pillars to cope with the additional loads for seating the lifting jacks that are needed to elevate the deck.|
Our findings indicate that decision-makers used options logic intuitively as a means to compare and contrast the alternative design solutions for the new viaduct. On the one hand, the urgency to replace the viaduct ruled out any solution that would immediately require elevating the deck of the viaduct. Even building just the viaduct with the deck at a higher elevation without elevating the highway would require building higher embankments at both ends of the viaduct. This, in turn, would involve purchasing land outside the current highway boundary for which the LA would need to initiate a public inquiry and statutory procedures that would likely prolong the job by up to four years. The LA reckoned that even if only the viaduct was elevated, adjacent highway segments would also need to be elevated at an estimated cost between £6 million and £7 million. On the other hand, aesthetic considerations and the rigidity of the culvert-based viaduct weighted strongly against such a solution, whose attractiveness stemmed from being fast to execute.
Our findings suggest that intuitive use of options logic ultimately informed the decision to build a multi-span viaduct with the option to elevate the deck at a reasonable cost in the future. Decision-makers agreed that building in this flexibility was worth the extra cost. Technically, this solution required more design work to guarantee the structural soundness of the deck. We did not uncover, however, any evidence that suggests the project team attempted to quantify the difference in capital costs between a rigid and a flexible multi-span viaduct. We also did not uncover data comparing the costs of exercising the deck-elevating option against the costs of demolishing and building a new viaduct in the future. Rather, the fieldwork suggests that the urgency to replace the old viaduct, and the marginal costs of flexibility compared with its potential benefits, ruled out any quantitative analysis irrespective of the technical capabilities of the design team. As the principal engineer and project manager explained:
“There was nothing scientific at all [in this decision]. A preliminary design for a flood-free route identified an appropriate gradient, but the scheme needed land take, which couldn't be funded. This was an emergency scheme, but its design allowed for raising the deck in the future. Additional design and construction costs were minimal, and no cost-benefit calculations were required.”
To investigate the effort that practitioners would need to incur had they opted to undertake a rigorous analytical assessment of this problem, Biesek and Gil (2010) modelled the two alternatives (with and without built-in flexibility). To this purpose, they used expanded NPV analysis that allows taking into account the value of flexibility built in upfront (Figure 1). Although the option to elevate the deck is not technically complex to design in, Biesek and Gil's (2010) study shows that modelling the decision analytically is not trivial. Expert assessments suggest that the cost of building a new viaduct with a built-in option would be around £3.5 million3 (including an allowance between 5% and 10% to build in the option), whereas the cost of elevating the deck alone in the future would be around £0.5 million. Alternatively, a cheaper but rigid viaduct could be built at around £3 million, which would need to be demolished (around £200,000) and replaced anew if the highway were to be elevated. Although assessing the construction costs of the two alternatives is relatively straightforward, analytical comparison of the alternatives also requires assessing the benefits from each one. This in turn requires assessing: (1) the value that the highway brings to the local economy, (2) the number of days that the highway would need to be closed if it were to be elevated as a function of the chosen design for the viaduct; and (3) when the highway would be elevated.
Figure 1: Investment outlay for the viaduct with and without option built-in.
Interestingly, Biesek and Gil's (2010) study shows that analytically modelling a seemingly straightforward design decision may actually require a large number of assumptions. The option to elevate the deck can be framed as a switch option (Smit & Trigeorgis, 2004) which is open in perpetuity. Mathematically, the problem can then be modelled using binomial or lattice valuation, which allow for a range of plausible process outcomes developing from one starting point (Cox, Ross, & Rubinstein, 1979). Although the mathematical apparatus itself is not overwhelming for experts (Copeland & Tufano, 2004), making the assumptions can be; specifically, it requires making assumptions on: (1) a risk-free rate of return; (2) an expiration date, for example, 120 years, (3) the stochastic fluctuation upward and downward of the value of the underlying asset over time, using for example a Geometric Brownian Motion (GBM) to model a ‘random walk’ (Cox et al., 1979)4, and (4) the relative value (Smit & Trigeorgis, 2004) between volatility in the value of the highway and in the costs of road closures, based on road traffic and closure statistics. Critically, it also requires assuming that the LA's decision as to when to switch will be based strictly on economic judgment (maximize the payoff) irrespective of local politics and budget constraints. Without any claim of prognostication, preliminary calculations indicate an overall net value of the option around £10 million (See Appendix I for details).
The analytical insight substantiates the decision taken by the LA to invest upfront in a design with built-in flexibility. Admittedly, the insight is contingent on various numerical and logical assumptions needed to produce a tractable model. To improve the quality of the insight, additional effort would be required to calibrate the assumptions and logic of a model that may not necessarily be replicated elsewhere. As a result, it would be unfair to criticize the consultant or the LA (both invariably operating on tight budgets) for not having invested in a formal evaluation of the alternatives. All in all, this suggests that using real options pricing models for supporting mundane design decisions may be overkill in a context of resource scarcity.
Notwithstanding this, this exploratory study shows that a real options cognitive lens is useful to frame design decision-making at front-end strategizing. The study reveals that a formal, as opposed to an intuitive, use of a real options lens can help to uncover variables and logic useful to assess the trade-offs between alternatives. This can help to make sensible and informed design decisions under conditions of uncertainty. We next discuss how this argument gets amplified as we study early design decision-making in multi-stakeholder project environments.
DESIGN DECISION-MAKING IN MULTI-STAKEHOLDER ENVIRONMENTS: THE CASE OF NR FRONT-END STRATEGIZING PRACTICES
NR, a company limited by public guarantee,5 came to existence in 2001, when the UK government decided to replace the commercially unsuccessful privately owned company, Railtrack. The firm owns, operates, maintains, and develops the main rail network in Great Britain6. Its mission is to ‘provide a safe, reliable, and efficient railway fit for the 21st century.’ NR's capital investment programme for 2009–2010 through 2013–2014 (Control Period 4) estimates an overall expenditure of £34 billion, splitting the budget between operations, maintenance, renewals, and enhancements teams. Between April 2007 and March 2009, for example, NR carried out £2.4 billion worth of enhancement work, including lengthening existing platforms or adding new ones, laying new track, and adding capacity through major re-signalling schemes. As with most capital organizations, the project development process at NR has been formalised. NR's Guide to Railway Investment Projects (GRIP) details the overarching sequence of project stages and respective deliverables (Table 2).
Table 2: NR's development process for capital projects (based on GRIP).
|#||Pre-GRIP||Initial planning and preparation to validate the project|
|1||Output Definition||Identify what the outputs will be and how they may be achieved|
|2||Pre-Feasibility||Detail the strategy of how to deliver the project outputs|
|3||Option Selection||Examine different options and select a single option to be developed|
|4||Single Option Development||Develop a single option at a high level and initiate the tendering process|
|5||Detailed Design||Award contracts and develop a detailed design and implementation plan|
|6||Construction, Testing, and Commissioning||Carry on physical works, ending with completion/commissioning|
|7||Scheme Hand Back||Hand back the asset to the asset owner, operator, or maintainer|
|8||Project Closeout||Update, finalise, and archive all the project documentation and capture the lessons learned|
|#||Post-GRIP||Demonstrate that the project has delivered its benefits|
Early Design Decision-making in a Multi-stakeholder Environment
To investigate the nature of early design decision-making at NR, we looked to front-end strategizing practices at three capital projects. We built a diverse sample by looking at two medium-size projects and a large one and by varying the project motivation (e.g., local socio-economic development, looming closure of an existing station) (Table 3). Unsurprisingly, we found that front-end strategizing invariably unfolds in a multi-stakeholder environment, where the number of stakeholders with a legitimate voice varies across projects. This adds a layer of complexity to early design decision-making, because achieving multilateral agreements often becomes a prerequisite to making decisions. Decision-making can be particularly challenging whenever; first, project funding is distributed across various pots, each one overseen by a different stakeholder, as opposed to coming from a single pot; second, organizations advocating particular long-term requirements cannot fund them; and third, the stakeholders’ visions are difficult to coalesce into a shared concept encompassing both short- and long-term design considerations.
Table 3: Context for design decision-making at front-end strategizing.
|Project||Arpley Chord||Reading||Salford Crescent|
|Description||Build a rail bypass to connect disparate lines and eliminate time-consuming manoeuvres||Build new platforms and concourses and improve layout to increase capacity||Improve layout of the station platforms to reduce overcrowding|
|Anticipated Final Cost (AFC)||~£15m||~£150m for the station; whole project costs over £650m||~£12m|
|Elapsed time for project life cycle||6 years (2008–2014)||9 years (2006–2015)||5 years (2009–2014)|
|Elapsed time for front-end strategizing (*)||~2 years (Oct 2008–Dec 2010)||~2 years (Feb 2006–Jan 2008)||~1.5 years (May 2009–Sept 2010)|
|Number of key stakeholders in frontend strategizing||Low |
e.g., Department for Transport (DfT), Council, NR, design consultant
e.g., University, station operator, NR, Council, Highways agency, public agencies
|Heterogeneity across stakeholders’ interests||High |
“a lot of our clients are not railway clients, they don't understand how the railway works, but we have to ensure that the projects don't infringe our ability to run the railway” [NR project manager]
“it took a lot of effort from us [to eliminate an unfeasible concept] and made us look pretty poor really” [NR programme manager explaining the difficulties in opposing council's ideas]
“the rail industry is quite clearly incredibly complicated and bureaucratic. And it's got understandably a lot of constraints and working methods” [Central Salford Urban Regeneration Company (URC) representative]
|Distribution of funding||One funding pot, potentially two||Major funding from DfT; other potential funding pots||Funding from earmarked DfT scheme; other potential funding pots|
|Sense of urgency||Low |
Project aimed at triggering local socio-economic development
Need to resolve major operational bottleneck and stop NR paying penalty fees
Closure of station imminent in 5 years unless redevelopment goes ahead
(*) Includes GRIP stages 1 to 3
The case of Arpley Chord, a third-party project that technically was not very complex, is telling. This £15 million project aimed to provide a new railway chord to connect disparate lines. The new chord would increase the operational flexibility and the capacity for rail traffic, eliminating the current need for inefficient run-round and turn-back manoeuvres7. The new layout would make redundant the existing assets, which once demolished could release land to build roads and property not far from the city centre. The project client, the Warrington Borough Council (WBC), a town with approximately 200,000 inhabitants, secured financial support for the scheme from a regional development agency8 as part of a regeneration plan. WBC's lack of expertise in railway projects made it imperative for the supplier to help WBC make adequate decisions at project front-end strategizing:
“It's our job to present the information unbiased. We've to clearly say: ‘look, these are your options’. And we try to present the information to enable them [WBC] to make their own decisions. This is important because, although they rely quite heavily on our advice, ultimately the accountability is with the funder” [NR Project Manager]
Likewise, we observed that early design decision-making for Reading Station—a £150 million project to redevelop one of the busiest railway stations in the United Kingdom with around 17 million annual users—also unfolded in a complex context of heterogeneous stakeholders’ priorities. In the short-term, the project was critical to resolve the severe capacity constraints that were making NR pay heavy fines to train operators (over £13 million in penalties in 2005–2006). Long-term considerations also informed the initial scope, which included building five new platforms, a new entrance, and an elevated railway to allow express trains to travel on fast lines over the slower lines, as well as lengthening three existing platforms. A prerequisite to design decision-making was forging multilateral agreements between NR, the DfT, and the Reading Borough Council. From the onset, the Council expressed interest in seeing a “world class 21st century station” that would catalyse local socio-economical development, but had limited resources. NR in turn was primarily interested in meeting DfT's targets for reliability, capacity, and safety using the fixed budget awarded by DfT. To succeed, the different stakeholders needed to coalesce their interests into a common vision that would fit with the available funding.
Interestingly, our findings reveal that the complexity of the institutional environment wrapping design decision-making can get amplified even in smaller projects, like the redevelopment of the critically overcrowded Salford Crescent Station. Although this project was urgent, because the lack of space for circulation at the platform posed serious risks to passenger safety, front-end strategizing lasted almost two years. The design concept needed to reconcile criteria in a DfT-sponsored programme providing £150 million of funding to support short-term improvements (National Stations Improvement Programme–NSIP) with NR's own long-term vision (encapsulated in its Route Utilisation Strategy–RUS–for the North West), and other stakeholders’ short- and long-term visions. Northern Rail, the franchised station operator, had a short-term interest in allowing its client First TransPenine Express operate with six-car trains (the station only accommodated less profitable five-car trains); it also had a long-term interest in adding a third platform. The University of Salford, whose campus is located on both sides of the station and is its main end-user, was interested in aligning upfront the project scope with its own £100 million capital programme to expand the campus over the next 20 years. In addition, a politically influential group of public agencies9 felt the project could be used instrumentally to advance their own agendas for urban regeneration and sustainable ways of commuting but was reluctant to use their own funds for this purpose. The project manager was not oblivious to this challenge: “we don't want to come up with a design for putting a really basic ticket office and if you want to improve it in the future, you'll realise you cannot do it because of the way we do things now.” Not surprisingly, reconciling multiple and occasionally conflicting needs (typically, poorly documented) in a design concept at front-end strategizing is not trivial. Next, we analyse the extent to which these multi-stakeholder project teams use options logic intuitively to support decision-making.
The Intuitive Use of Options Logic in Early Design Decision-making
The analysis of our findings systematically indicates that multi-stakeholder project teams resort intuitively to options logic to frame design decision-making problems at front-end strategizing. We also find that decision-makers repeatedly need to factor in multiple options, which compete among themselves for a limited if not fixed pot of funding—a problem hard to model using real options (Driouchi, Leseure, & Bennett, 2009). Interestingly, we show that the options themselves are frequently not technically complex to design in. Rather, they may simply require mundane investments in design provisions to facilitate future growth, stage project delivery, or build operational flexibility. But because the options compete with one another for resources, they require decision-makers to collectively agree to make trade-offs under uncertainty and conflicting visions about the future. As one respondent put it: “it's amongst the easiest things to identify what we might do to ‘future-proofing'; the hardest is to say who is gonna pay for that?” We observed significant variability across the motivations for designing in the options, who advocates them, and the extent to which asymmetries in knowledge or in decision-making power exist across the different parties. Some options are advocated by the designers and purely motivated on technical grounds, whereas others are advocated by public agencies advancing socio-economic development agendas. Importantly, the advocator may not necessarily be in a position to fund the option.
Overall, our findings suggest that the collective design decision-making process systematically unfolds mostly unaided by any sort of formal frameworks. As one respondent put it, “I future-proof. But that's me doing what I believe is right. I believe in railways and in the service they provide. But that doesn't play into the corporate vision, and perhaps I'm wasting company money.” Others noted that the informal, hazardous way to make decisions could also reward those ‘that get it wrong,’ failing to plan for the future but saving money in the present. The quality of the conversation and outcome is thus dependent on decision-makers’ personalities and the information people happen to bring to meetings, knowledge asymmetries, and intuitive options logic as opposed to a structured debate under uncertainty.
Table 4: Framing early design decision-making in terms of options logic.
|Project||Arpley Chord||Reading||Salford Crescent|
|Sample of Options observed||(1) Electrify the railway line; |
(2) Increase rail gauge from W10 to W12 10
|(1) Increase capacity of the station; |
(2) Connect new station to private development
|(1) Add third platform; |
(2) Build landmark station building
|Types of options||Switch operations regime (1,2)||Grow capacity (1) stage delivery (2)||Grow capacity (1,2)|
|Perceived value of the options||Reduce costs for modernizing the line in the future (1,2)||Reduce costs for further modernizing the station in the future (1,2)||Reduce costs for further modernizing the station in the future (1,2)|
|Advocator of the option||NR technical division (1,2)||Council (1,2)||Multiple stakeholders (1,2)|
|Assessment of option cost||Additional £0.5m to the construction costs||~£1m (less than 1% of the whole budget)||Need for additional land and more expensive station building|
|Assessment of options value||Intuitive logic||Rudimentary cost-benefit analysis||Intuitive logic|
|Funding||Two pots of funding||Options affordable within DfT's project funding||Multiple pots of funding|
The Arpley Chord case provides a good illustration of the intuitive use of options logic at early design decision-making. Because the layout of the new chord would conflict with an existing road, a new viaduct over the railway needed to be built. From the onset, the engineering team aimed to ‘future-proof’ the viaduct by designing in volumetric clearance to avoid demolishing it if NR wanted to change the railway line to a larger gauge (a prerequisite to increasing capacity) and/or electrify the line in the future (a change necessary to lowering CO2 emissions). As the senior route planner put it:
“Any new chord designed and built MUST be clear for W12 gauge and full electrification clearance. It would not survive any network change consultation if not”
The team expected that sizing the viaduct for a long-term scenario would add around £0.5 million to the construction costs. Interestingly, the question as to whether it should be the Council funding these options as opposed to NR because they were needed anyway to meet NR's design standards has never emerged. One could argue that the investments were aligned with the Council's long-term regeneration agenda, and ultimately it was about public money one way or another. But the findings also reveal a deep asymmetry in technical knowledge between NR and its client. Although NR's project manager argued that “it was their job to present the information unbiased,” another would be quick to say that he always tried to “future-proof to the best of his ability.” They would acknowledge they lacked guidance as to whether electrification clearance should be built in, and in the case of viaducts, future-proofing was often a “wonderful utopia since there could be over 100 miles of low bridges and tunnels.”
The redevelopment of the Reading Station is another example of intuitive use of options logic to frame early design decision-making problems, and of ad hoc asymmetries determining whether options became framed as ‘nice to have’ as opposed to ‘have to have.’ After NR was awarded funding for the project by the DfT, the Council became interested in seeing a commitment in the planning application to connecting the station to the Station Hill—a £400 million private development in a five-acre area adjacent to the station. To this purpose, the council asked NR to design in options to connect the station to the potential new development and to allow further expansion of the concourse in the future. No private developer had yet committed, however, to making the required investment, and the financial crisis made it difficult at the time (and still makes it difficult) to predict how things will pan out. Together, the two options would add over £1 million to the capital investment, which NR had not factored in when it bid for DfT money to redevelop the station. Our findings were inconclusive as to whether NR consulted the Council properly when it put together the bid. But they show that the decision to design in the two options after the project was awarded DfT money was protracted. Although the council insisted in designing in the two options, it had no means to fund them. NR, in turn, operated with a fixed budget and only built in the options as it found savings in other parts of the project. Importantly, our findings suggest that the outcome hinged in part on the personality of the Council's representative, whose support could be perceived as instrumental to NR for getting planning consent. As the NR programme manager put it:
“This Council person, a very strong character, she was probably the reason why these ideas went forward. I mean, NR's view is to do what we have been asked to. As an organization, we don't really care about the streets of Reading. And DfT, well, that's not their game.”
In contrast, the case of Salford Crescent reveals an environment where options logic was instrumental to resolving the design in the face of negligible asymmetries in knowledge or decision-making powers across stakeholders. Two fundamental options surfaced early on: adding a third platform and a landmark station building in the future. There was high uncertainty particularly as to whether a costly third platform (around £30 million) will ever be added to the station. This option was aligned with the strategic visions of NR to grow capacity, even if a NR respondent asked rhetorically: ‘How many of us really read RUS? It's being updated on-line, but do you understand it?” The option was also aligned with Northern Rail's long-term plans to grow capacity, the University's plan to grow its campus, and the public agencies’ interests in further socio-economic development. For example, representatives of Central Salford URC asserted:
“In ten years’ time, this Station will be completely different because the university is going to grow, and they are changing their estate. The roads that run pass the university are also going to be developed. We're spending something in the region of £40 million at the moment!”
But, at the same time, high uncertainty existed around the materialisation of these visions in an age of austerity. As an NR representative put it: “the trick is working out a realistic scenario in ten years’ time. What wasn't realistic in the past might be realistic now, and some comes down to crystal gazing your assumptions.” The university, for example, could not pin down reliable projections for growth in student numbers because undergraduate fees were about to treble, whereas the other public agencies were struggling financially in the aftermath of the financial crisis. As a result, these interested parties shied away from committing any funds to the options. In turn, the NISP government scheme, the main mechanism used to fund the project, excluded investments in new platforms or in modernizing the railway line. Delays in reaching an agreement were exasperating to the extent that the NR commercial sponsor warned in one meeting that ‘if you don't come along with the funding, sooner or later, we'll have to put something in there.’ Eventually, the decision-makers agreed to use the DfT's funding pot to design in limited safeguards (Gil, 2007) for a third platform and landmark building.
In summary, our findings reveal multi-stakeholder teams systematically using options logic to support early design decision-making. But they do it intuitively at best and often in the midst of extreme uncertainty in requirements, conflicting visions, and sharp asymmetries in knowledge and decision-making powers, which arguably can lead to issues of fairness and ethics in stakeholder management (Phillips, 2003). Insights from a workshop further pointed to an opportunity to formalise options logic. As one participant noted:
“We have to have a basic understanding of which direction we want to go in. I think this is what we are talking about with evolvability. It's almost like when we brought in value management – we always did it unscientifically and then became more formalised.... I think this [design for evolvability] is about a structure behind asking the question high up, before you start off doing all this - is there something simple you could do to future-proof?”
We next present a method to formalise this process in a multi-stakeholder context, which resulted from cross-fertilizing our empirical findings with theory in real options reasoning.
DESIGN FOR EVOLVABILITY: PROPOSED METHOD
The method to design for evolvability consists of a structured process to formalise the use options logic in concept design at the project front-end strategizing. The aim is to ensure the project team follows a systematic approach to decide whether provisions to cope with foreseeable uncertainties in requirements should be incorporated in the upfront design of a new asset with an expected long operational life. To this purpose, the developer appoints a Champion of Design for Evolvability, who is empowered to steer the legitimate stakeholders to agree on a concept that factors in both concerns of short-term affordability and long-term adaptability. The method involves three sequential stages, allowing for iterative loops: (1) analysing options, (2) designing alternatives, and (3) project strategizing. Each stage, in turn, encompasses a sequence of steps aimed at producing a deliverable that feeds into the next stage (Figure 2). A detailed explanation of the method follows.
Figure 2: Schematic representation of the design for evolvability method.
Analysing Options (Stage 1)
The first step in the analysing options stage is Identify Potential Options. The champion of design for evolvability will first ask the project stakeholders to check whether their parent organisations have developed a strategic vision, and, second, the extent to which any eventual strategic plans translate into a range of possible operating scenarios for the new asset in the future. These scenarios should inform in turn the identification of the potential options that can be built into the design of the new asset in light of foreseeable uncertainties in functional and operational requirements. The stakeholders will also be asked to investigate whether their organisations already have lists of built-in options that may have been developed for comparable projects in the past. Although capital projects tend to be one-off, many similarities tend to be found at the development process level from project to project. Over time, the stakeholders can be expected to have built a repository of knowledge that captures design uncertainties that recur systematically, which is useful to inform a conversation on design for evolvability.
Once the options have been identified, the project teams need to qualify options. In options logic, the expiration date is the date after which an option can no longer be exercised, whereas the exercise date defines the expected timescale for undertaking the right to exercise the option; some options stay open in perpetuity. Often, decision-makers only find attractive investments in options likely to be exercised in the near future because of the shorter payback period (Gil, 2007). It is critical that the champion of design for evolvability shares his or her expertise at this stage because the option value actually increases with long expiration dates, as it allows more time for uncertainties to resolve favourably and thereby for decision-makers to exercise the option in light of new information. To qualify the options, the project stakeholders will be asked to characterise: (1) the value to be created if uncertainty resolves favourably; (2) the expected timescale for exercising the option and whether the option eventually expires at some point in the future; and (3) the likelihood that each option will be exercised.
To conclude this stage, the project stakeholders will be asked to characterize option sponsorship. The champion of design for evolvability is responsible for starting a conversation about which organizations are in a position to fund the capital costs (or part of) that need to be incurred to build in each option. The output of this stage is a deliverable that documents the rationale underpinning each potential option as well as option sponsorship and respective funding commitments (Table 5).
Table 5: Stage 1 deliverable: Qualified options.
|Option A||“insert title of potential option here”|
|Foreseeable uncertainty||“spell out the motivation for building this option”|
|Design requirements||“describe the range of design requirements that might be needed”|
|Estimated additional value||“estimate the expected range for the value of the option”|
|Expected timescales||“spell out when this option may be expected to be exercised”|
|Potential expiration Date||“indicate date when the option can no longer be exercised (if existent)|
|Likelihood of being exercised||“spell out the likelihood of exercising the option in the above timescales”|
|Option Sponsorship||“identify key sponsors and additional sponsors and their funding commitments|
Designing Alternatives (Stage 2)
In the first step of the second stage, project teams will be asked to identify alternative concepts that vary to the extent in which they build in the design the options identified in the previous stage, and accordingly vary in the costs to be incurred for building the options and for exercising them in the future. The project teams will also be asked to characterise the key features of each concept. Concepts that may have surfaced in previous projects should be considered— there is no need to reinvent the wheel for every new project. Importantly, the purpose is not to discuss alternative design solutions to address immediate needs, —an exercise often performed through the so-called ‘optioneering appraisals.’ Rather, the purpose is to discuss alternative ways to build in options to cope with foreseeable change in the future. Particular attention should be given to whether design architectures can be modularised, as by definition, modular designs have options built-in (Baldwin & Clark, 2000). Alternatively, if modularity is difficult to achieve, options can be built in integral architectures through investments in safeguards (e.g., over-sizing foundations, increasing floor-to-ceiling heights, and conservative equipment choices). (Gil, 2007) Teams will also be asked to consider a baseline scenario that rules out investments to build in options.
Once the design concepts have been identified, project teams will be asked to assess the costs of each alternative in terms of the capital costs to build in the options and the expected costs of exercising the options in the future. Teams should assess the costs of adaptation for the baseline scenario (without built-in options), assuming this scenario does not prohibit adaptation (Table 6). The champion of design for evolvability will be tasked with educating the project team about the implications of ruling in and out potential options, using examples for the sake of illustration. For instance, safeguarding the option to add more storeys to a building by over-sizing its foundations and columns increases the construction costs, but it reduces the cost of adding the storeys and disrupting operations in the future. Alternatively, the project team may choose to make a more moderate investment that does not prohibit future modifications, for example, over-sizing the foundations but not the columns. The latter is marginally more affordable initially, but it will be more costly to adapt the building and certainly a lot more disruptive to operations in the future. For the baseline scenario (e.g., don't even over-size the foundations) adaptation may be eventually prohibitively costly, which represents a risk of obsolescence if uncertainties resolve favourably. Often doing something small now prevents much more trouble later, provided the investor can afford the upfront commitments.
Table 6: Stage 2 deliverable: Qualified alternatives.
|Option A||“insert title of potential option here”|
|Alternative concepts||Baseline Scenario: No option built-in||Alternative 1 : Option partially built-in||Alternative 2: Option fully built-in|
|Additional capital costs||0|
Mindful that estimating the costs of option exercising might be challenging as projections need to be made into the future, the champion of design for evolvability should ask teams to estimate ranges of values as opposed to producing single-point estimates. As the teams pin down information on the costs, they should double check whether the option sponsor(s) can indeed commit to funding the capital costs.
To conclude this stage, project teams will be asked to document critical assumptions made for qualifying the options and alternative concepts. This step is important to enable the project team to revise the validity of these assumptions over time. For example, radical changes in the external environment may significantly affect the costs and value of an option to the extent it may lead to (1) abandoning options initially thought to be worth investing in, or (2) building in an option previously ruled out from the design for evolvability strategy.
Project Strategizing (Stage 3)
In the final stage, the project team will be expected to recommend a design for evolvability strategy as part of the front-end strategizing effort. The team will have to agree collectively whether they recommend building in any specific options through modular or safeguarded architectures, and if so, the level of capital investment required. Alternatively, they can recommend ruling out any investment in built-in options. The strategy should spell out the capital costs involved, the costs of adapting the asset if foreseeable uncertainties are realised in a favourable way in the future, and option sponsorship. In addition, the project team will also be asked to specify design for evolvability checkpoints when the team will be expected to check the validity of the assumptions informing the strategy. The checkpoints need to be set up in a way that fits with the design and development process adopted by the leading party.
We acknowledge the financial support of Project Management Institute and Manchester Business School. We also acknowledge the contribution in kind of Network Rail. The arguments developed here result from the analysis we undertook and do not express the opinions of the respondents.
Amram, M., & Kulatilaka, N. (1999). Real options: Managing strategic investment in an uncertain world (p. 246). Boston: Harvard Business School Press.
Baldwin, C. Y., & Clark, K. B. (2000). Design rules: The power of modularity (Vol. 1). Cambridge, MA: MIT Press.
Barnett, M. L. (2005). Paying attention to real options. R&D Management, 35(1), 61–72.
Biernacki, P., & Waldorf, D. (1981). Snowball sampling: Problems and techniques of chain referral sampling. Sociological Methods & Research, 10(2), 141. SAGE Publications. doi:10.1177/004912418101000205
Biesek, G., & Gil, N. (2010). Project design for evolvability: Using options logic at front-end strategizing.
Bowman, E. H., & Moskowitz, G. T. (2001). Real options analysis and strategic decision making. Organization Science, 12(6), 772–777.
British Standards (1988). BS 5400-1:1988 - Steel, concrete and composite bridges. General statement. (B. Standard, Ed.).
Browning, T. R. (2001). Applying the design structure matrix to system decomposition and integration problems: A review and new directions. IEEE Transactions on Engineering Management, 48(3), 292–306.
Brusoni, S., & Prencipe, A. (2006). Making design rules: A multidomain perspective. Organization Science, 17(2), 179–189.
Chapman, C., & Ward, S. (1998). Project risk management: Processes, techniques and insights. Chichester, UK: John Wiley & Sons.
Cleland, D. I., & King, W. R. (1983). Systems analysis and project management. New York: McGraw-Hill.
Cooper, D., & Chapman, C. (1987). Risk analysis for large projects: Models, methods, and cases. Chichester, UK: Wiley, John & Sons, Incorporated.
Copeland, T., & Tufano, P. (2004). A real-world way to manage real options. Harvard Business Review. Retrieved from http://cours2.fsa.ulaval.ca/cours/gsf-64901/HBR_Copeland_March 2004.pdf
Cox, J., Ross, S., & Rubinstein, M. (1979). Option pricing: A simplified approach. Journal of financial Economics, 7(3), 229–263. doi:10.1016/0304-405X(79)90015-1
Dixit, A. K., & Pindyck, R. S. (1994). Investment under uncertainty (p. 468). Princeton: Princeton University Press.
Driouchi, T., Leseure, M., & Bennett, D. (2009). A robustness framework for monitoring real options under uncertainty. Omega, 37(3), 698–710. doi:10.1016/j.omega.2007.09.004
Eppinger, S. D., Whitney, D. E., Smith, R. P., & Gebala, D. A. (1994). A model-based method for organizing tasks in product development. Research in Engineering Design, 6(1), 1–13.
Ethiraj, S. K., & Levinthal, D. (2004). Modularity and innovation in complex systems. Management Science, 50(2), 159–173.
Fixson, S. K., & Park, J.-K. (2008). The power of integrality: Linkages between product architecture, innovation, and industry structure. Research Policy, 37(8), 1296–1316.
Ford, D. N., Lander, D. M., & Voyer, J. J. (2002). A real options approach to valuing strategic flexibility in uncertain construction projects. Construction Management and Economics, 20(4), 343–351.
Gil, N. (2007). On the value of project safeguards: Embedding real options in complex products and systems. Research Policy, 36, 980–999.
Gil, N., & Tether, B. (2011). Design flexibility and risk management: Substitutes or complements? Insights from Heathrow's Terminal 5 Project.
Gil, N., Tommelein, I. D., & Schruben, L. W. (2006). External change in large engineering design projects: The role of the client. IEEE Transactions on Engineering Management, 53(3), 426-439. doi:10.1109/TEM.2006.877447
Gil, N., Tommelein, I. D., Stout, A., & Garrett, T. (2005). Embodying product and process flexibility to cope with challenging project deliveries. Journal of Construction engineering and management, 131(April), 439. Retrieved from http://link.aip.org/link/?JCEMD4/131/439/1
Highways Agency (1992). BD 36/92 Design manual for roads and bridges. (T. H. Agency, Ed.).
Kalligeros, K. (2006). Platforms and real options in large-scale engineering systems. Engineering Systems Division. Massachusetts Institute of Technology, Massachusetts.
Lander, D. M., & Pinches, G. E. (1998). Challenges to the practical implementation of modeling and valuing real options. The Quarterly Review of Economics and Finance, 38(Special Issue), 537–567.
Lee, Y. S. (2007). Flexible design in public private partnerships: A PFI case study in the National Health Service. University of Cambridge.
Lessard, D., & Miller, R. (2000). Mapping and facing the landscape of risks. The strategic management of large engineering projects. Cambridge, MA: The MIT Press.
MacMillan, I. C., Putten, A. B. van, McGrath, R. G., & Thompson, J. D. (2006). Using real options discipline for highly uncertain technology investments. Research-Technology Management, 49(1), 29–37. Industrial Research Institute, Inc. Retrieved from http://www.ingentaconnect.com/content/iri/rtm/2006/00000049/00000001/art00005
McGrath, R. G. (1997). A real options logic for initiating technology positioning investments. The Academy of Management Review, 22(4), 974–996.
McGrath, R. G., & MacMillan, I. C. (2000). Assessing technology projects using real options reasoning. Research Technology Management, 43(4), 35–49.
Miller, R., & Lessard, D. (2000). The strategic management of large engineering projects. Cambridge, MA: The MIT Press.
Miller, R., & Lessard, D. (2001). Understanding and managing risks in large engineering projects. International Journal of Project Management, 19(8), 437–443.
Miller, R., & Lessard, D. (2007). Evolving strategy: Risk management and the shaping of large engineering projects. MIT Sloan School of Management.
Morris, P. W. G. (1994). The management of projects. London: Thomas Telford.
Neufville, R. de, & Scholtes, S. (2011). Flexibility in engineering design. MIT Press.
Phillips, R. (2003). Stakeholder theory and organizational ethics. Berrett-Koehler Pub.
Sanchez, R., & Mahoney, J. (1996). Modularity, flexibility, and knowledge management in product and organization design.pdf. Strategic Management Journal, 17(winter), 63–76.
Schilling, M. A. (2000). Toward a general modular systems theory and its application to interfirm product modularity. The Academy of Management Review, 25(2), 312–334.
Simon, H. A. (1962). The architecture of complexity. Proceedings of the American Philosophical Society (Vol. 106).
Smit, H. T. J., & Trigeorgis, L. (2004). Strategic investment: Real options and games. Oxfordshire, UK: Princeton University Press.
Sreeves, J. (2007). Future-proof: Upton-upon-Severn viaduct, UK. In T. T. S. LTDA (Ed.), Civil Engineering (pp. 33–38). ICE.
Steward, D. V. (1981). The design structure system: A method for managing the design of complex systems. IEEE Transactions on Engineering Management, 28(3), 71–74.
Suh, E. S., Weck, O. L. de, & Chang, D. (2007). Flexible products platform: Framework and case study. Research in Engineering Design, 18, 67–89.
Thomke, S. H. (1997). The role of flexibility in the development of new products: An empirical study. Research Policy, 26(1), 105–119.
Trigeorgis, L. (1996). Real options, managerial flexibility and strategy in resource allocation. Cambridge, MA: MIT Press.
Ulrich, K. (1995). The role of product architecture in the manufacturing firm. Research Policy, 24(3), 21.
Wang, T. (2005). Real options “in” projects and systems design identification of options and solutions for path dependency. Engineering Systems Division. Massachusetts Institute of Technology.
Zhao, T., Sundararajan, S. K., & Tseng, C.-L. (2004). Highway development decision-making under uncertainty: A real options approach. Journal of Infrastructure Systems, 10(1), 23-32.
APPENDIX I: Using Real Options to Make Sense of the Upton-upon-Severn Viaduct Design
To make sense of the design of this viaduct, we framed the possibility of elevating the deck in the future as a switch option (Smit & Trigeorgis, 2004) and used binomial or lattice valuation to replicate the characteristics of the process (Cox et al., 1979). To estimate the range of possible values for the viaduct, we assume a current value V0 and divide time in Δt increments. The value of the asset can increase by a factor u (Vu,t=uV0), or decrease by a factor d (Vd,t=dV0); and p represents the probability of V0 performing an upward movement and 1-p the probability of V0 performing a downward movement. The variables u, d, and p can be calculated as follows:
where σ is volatility in project value over time and r is the risk-free rate of return.
Based on the 2.4% volatility in the traffic flow obtained from the DfT road traffic statistics (DfT, 2007), assuming interest rates of 2%, and starting with V0=£134.4m11, we estimate how the viaduct value fluctuates over time until the expiration date (which we assumed to be 120 years). The fluctuation of costs of traffic interruptions is calculated in a similar manner, using 20% as the volatility (obtained from the flood events statistics) and starting point of C0= £2.28m. We limit the maximum traffic flow to the road capacity (18,000 vehicles per day), and the number of interruptions per year to 365 days. Figure A.1 depicts a plot of the event tree for the fluctuation of value and costs.
Figure A.1: Event tree representation for value and costs.
We compute the payoff P of having an option to switch at each end node, which is the maximum between the profits from the exercised option (V-Cexe) discounted by the profits from the current project (V-C) and 0. Whenever the profits from the exercised option are higher than those from the current project, having an option pays off, and investors will presumably switch (i.e., exercise the option)
where Cexe is the exercise cost (estimated at £500,000 for this case).
Working back from the payoff at the expiration date, we assess the value of having the opportunity to switch. The option value O at the any end node can then be expressed by:
where Pu is the upside payoff and Pd is the downside payoff
1 The expression future-proof is often used by the industry to refer to what we term design for evolvability.
2 The Upton-upon-Severn Viaduct project received the Construction Award and the Overall Project Award from the Institution of Civil Engineers (ICE) West Midlands in 2005.
3 All costs of this exploratory study are given in 2004 prices.
4 ‘Random walk’ applies if the asset value can be assumed to vary along a steady long-term trend similar to variations in prices for stocks and other publicly-traded assets (Cox et al. 1979).
5 A guarantee company does not usually have a share capital or shareholders. All profits are therefore available for re-investment in the rail network.
6 This includes around 20,000 miles of tracks; 40,000 bridges, tunnels, and viaducts; 6650 level crossings; and 2500 stations mainly leased to train operators.
7 The freight operator believed the new chord could save around 30 minutes of journey time, as well as improve working conditions and safety.
8 Northwest Regional Development Agency (NWDA), a public body set up to help improve the quality of life and economic prosperity of the North West region that is closing shop in 2012.
9 Central Salford URC, Salford City Council, and Greater Manchester Passenger Transport Executive
10 The rail gauge is the distance between the inner sides of the heads of the two load bearing rails. The loading gauge defines the maximum height and width for railway vehicles and their loads. NR uses a W loading gauge classification system that ranges from W6a (smallest) to W12 (largest).
11 V0 and C0 are estimated based on the value the highway brings to the local economy as well as the effects of road closure caused by floods. Details can be found in Biesek and Gil (2010).
©2012 Project Management Institute