Alliance Contracting—Evolving Alternative Project Delivery (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

40 chapter three ALLIANCE CONTRACTING PROCUREMENT POLICIES, PROCEDURES, AND PROGRAMS INTRODUCTION This chapter reviews findings as they relate to the policies, principles, and guidelines currently being followed by state transportation agencies to implement alliance contracts for infrastructure construction and maintenance projects. This chapter will combine information collected through the liter- ature search, the screening survey, and interviews with case example agencies and industry professionals. The focus of this chapter is the pre-award tasks necessary to bring an alli- ance into existence. ALLIANCING POLICIES Having a clear target of what the agency wants to produce is the key to establishing policies and procedures to implement any new form of project delivery. The “target” is defined essentially by the reason the agency chose to deviate from traditional forms of project delivery. Alliance contracting is no different. Often the motivation to deliver an alliancing project is a requirement for enhanced collaboration and integration during early proj- ect development phases to gain the benefits of constructability, biddability, and maintainability in the project’s design docu- ments (van den Berg and Kamminga 2006). One author pro- vides some insight on the target addressed by alliancing: An alliance can be defined as a working partnership in which there is mutual recognition and understanding that the success of each firm depends in part on the other firm/firms . . . The pooled advantages can stem from each organization’s strengths compensating for the other’s weaknesses or from amplifying or enhancing their combined strengths. (Iyer 2003) The alliance contracts are relational contracts and offer a fundamentally different approach to targeting outcome and sharing risk. The most well-used alliancing model is now com- monly known as the “pure” alliance. The literature is rich with analyses reporting the advantages and disadvantages of the pure alliance (Green 1999; Uher 1999; Bresnen and Marshall 2002; Naoum 2003). In the past few years, new variants of the alliance model have appeared with significant differences from the pure alliance model. Relational contracts of any type require “a change . . . within the modus operandi of the indus- try and its clients” (Lahdenperä 2012). Agencies in Australia and New Zealand regard them as “strategic means to improve the performance of their core operations and it is considered that these variants may offer a project delivery model that is suited to different situations” (Lahdenperä 2012). The new variants are termed “competitive alliancing” and “collabora- tive alliancing.” They are very much the same as pure allianc- ing, including the risk allocations and the project management structures, during design and construction. However, the “com- petitive” alliance requires multiple teams to compete for the award of the project and a collaborative alliance involves multi- ple members competing for work during the project based on their past performance in previous alliance work. The case studies showed that there were three primary areas in which policy had been developed to implement alliancing. 1. Project selection policy, 2. Policy for selecting the type of alliance, and 3. Policy for the selection of alliance members. Alliance Project Selection “Many experts believe that the key to the success of a con- struction project is the process by which it is organized and managed; i.e., the project delivery method” (Reicke 2004). Therefore, it is important to select a project whose character- istics make it a good candidate for alliance contract delivery. The 2009 NZTA Procurement Manual notes that large scale, high-risk projects in difficult environments with external stakeholder issues must be addressed during design and con- struction as having the project characteristics that lend them- selves to alliancing. Table 18 is a summary of the project characteristics that were found in the policy documentation from Australia, New Zealand, and the United Kingdom. Synthesizing the information contained in Table 18 by selecting characteristics with at least four observations leads to the conclusion that projects with the following listed set of characteristics are good candidates for alliance contracting delivery: • AU$50 million (~U.S. $47 million) or more: The proj- ect must be of sufficient size to make potential cost sav- ings as a result of collaboration and integration worth the cost of the additional upfront expenses to form the alliance for all parties to the agreement. • High-risk profile: “Alliancing will generally be appro- priate for projects that are characterized by major risks and complexities that cannot be dimensioned in the

41 ancing. The QDTMR Relational Contracting Guide contains a helpful tool for determining if a project is a good candidate for alliance contract delivery, which is shown in Table 19. Lastly, the Australian State of New South Wales Procure- ment Methodology Guidelines for Construction (2005) pro- vides its lessons learned regarding those projects that are not good alliance candidates. The following circumstances provide an indication of where alliancing is unsuitable: • The personnel involved (from the agency and other stake- holders, consultants and contractors) are not experienced at working together and unwilling or unable to adopt the atti- tudes and corporate cultures necessary to work as a team; • The agency is not prepared to invest the resources required to participate in a relationship contract and accept a risk sharing arrangement; • The project is relatively small, and the additional tendering and implementation costs are disproportionate with the cost of the work and the likely benefits; • The . . . budget and financial risks to . . . the state are too great to enter into a commercial arrangement where project costs or schedules are uncapped; or • More conventional contracts will achieve the outcomes required, since the project is not complex, risks are well understood or there is little room for improving outcomes or issues can be resolved without contractor involvement early in the design process. (NSW 2005) Once the agency has convinced itself that a given project will benefit from alliance contracting project delivery, it must then determine which type of alliance contract is most appro- priate for the specific project. Business Case or soon thereafter. This is likely to have a material but indeterminate impact on achieving project objectives, which means that the flexibility and collab- orative decision made under an alliance contract may be desirable. Under the alliance approach, the parties can deal with any risks and complexities if and when they arise over the life of the project” (ADIT 2011). • Highly complex: Complexity extends from merely the engineering challenges into attempting to control external contextual and financial factors that ultimately impact the final engineering solution. Therefore, a project needs to be one where having early contractor involve- ment in both planning and design would be expected to generate the optimal design with respect to social, politi- cal, environmental, and technical demands. • Large number of stakeholders: “. . . project characteris- tics that lend themselves to Project Alliancing are large scale, high-risk projects in difficult environments that have complex stakeholder issues that require flexibility during design and construction” (Le Masurier 2006). • Need for flexibility: In this context, flexibility is defined as the ability to make decisions and change plans in an expeditious and agile manner. It also speaks to the risk “potential for a substantial change in project scope” (Queensland 2008). • Aggressive schedule: “Tight time constraint is another significant reason for using alliancing” (Chen et al. 2012). The ranking of the project characteristics indicates that alliancing is used primarily as a risk management approach on large complex projects in the three countries that use alli- TABLE 18 ALLIANCE PROJECT CANDIDATE CHARACTERISTICS Nation UK Australia New Zealand TotalsProject Characteristic HA VDTF Austroads ADIT QDTMR NZTA Size — >AU$50 million >AU$50 million >AU$50 million >AU$30 million >NZ$100 million >AU$50 million* Risk Profile High High High High High High 6 Complexity High High High — High High 5 Number of Stakeholders High — High — High High 4 External Factors High — High — Special need High 4 Need for Flexibility — High High — High High 4 Schedule — Urgent start required Time pressure Urgent start required Limited — 4 Technical Requirements High High High — — High 3 Scope Definition — Uncertain Uncertain — Unclear — 3 Need for Innovation — — High High High — 3 Agency Resources Limited — — Insufficient — Insufficient 3 Competition — — — Low — Low 2 Demonstrate VfM High — — — — Low 2 Budget — — Tight — — — 1 *Average = AU$56 million. — = not applicable.

42 picked by the QBS process consist of a single alliance team. Competitive alliance partners are selected on a best-value basis with an emphasis on the proposed TOC; however, they do not compete for work after selection. Figure 15 illustrates NZTA’s comparison of the three alternatives against the dif- ferent phases in a project. Initially, an agency considering a collaborative alliance must develop the project’s business case and funding using its internal staff. Next, the agency proceeds to tender and selects Alliance-Type Selection When choosing the type of alliance to enter into it is useful to consider the progression of procurement actions each takes and how the various parties to the alliance are involved. The collaborative alliance model differs from pure and competi- tive alliancing in that it consists of different teams that have to collaborate but also compete for work during the project on a basis of the quality and timeliness of earlier work com- pleted in that alliance. Pure alliances completed with partners TABLE 19 QUEENSLAND ALLIANCE SELECTION DECISION MATRIX Question Alliance Selection Is the project value in excess of $30 million? Score 3 for Yes, 0 for No Can project risks be equitably assigned to contractors? Score 0 for Yes, 3 for No Are complex community and stakeholder issues expected? Score 1 for Yes, 0 for No Are there significant schedule constraints? Score 1 for Yes, 0 for No Does the procurement agency have the requisite resources to support an alliance board and provide input into the alliance? Score 2 for Yes, 0 for No Is there a requirement for flexibility in project delivery; e.g., potential for substantial change in project scope? Score 3 for Yes, 0 for No Are there a sufficient number of industry participants available to provide competitive responses to requests for tender? Score 0 for Yes, 1 for No Is the alliance owner capable of embarking on relational style contracts? Score 2 for Yes, 0 for No Are alliance participants vendors of equipment that is integral to the delivery of the project? Score 0 for Yes, 1 for No Is the project subject to high environmental or cultural risks? Score 1 for Yes, 0 for No Can high-risk elements of the project be separated from the main project? Score 0 for Yes, 1 for No Is the procurement agency willing to accept a commercial arrangement with uncapped costs and schedules? Score 1 for Yes, 0 for No Total Score 0–7 not suitable for alliancing; Score 8–14 consider using alliancing Score 15–20 alliancing is highly suitable Source: Queensland (2008). FIGURE 15 New Zealand alliance contracting options (Scheepbouwer and Gransberg 2014).

43 to minimize the probability job losses, which must be shared by all at the end of day. The juxtaposition of internal com- petition based on performance on the pain/gain share creates an environment that enhances collaboration and encourages the construction teams to provide outstanding outcomes in the KPIs and KRAs. The situation incentivizes all teams to perform at their best to secure a larger share of the work, and follow up by delivering it for best cost. Alliance Member Selection The literature review revealed a surprising number of refer- ences to selection based on individual personnel who had the correct skills, experiences, and, mostly importantly, the attitude that fosters collaboration. For example, the Victoria Project Alliancing Practitioner’s Guide (2006) directs that “an integrated project team [be] selected on the basis of best person for each position.” Chen et al. (2012) recommends that team selection criteria be based on: “Focus on partners’ competence, reputation and attitude; and select personnel on a ‘best for project’ basis.” The Austroads Building and Construction Procurement Guide (2011) cites “difficulty sourcing personnel with the right personal attributes and pre- paredness to work in an alliance structure” as a potential dis- advantage of alliancing. Another source advocates retaining “behavioral coaches” to assist the agency selection panel to “develop a better understanding of the Proponent’s [poten- tial partner] potential to form an alliance with the Owner’s personnel” (ADIT 2011). The same source goes on to rec- ommend that after the project alliance agreement is signed, “the alliance may decide to use a Behavioural [sic] Coach to consolidate alliancing behaviours [sic].” The Australian Department of Infrastructure and Transport advocates retaining “behavioral coaches” to assist the agency selection panel “develop a better understanding of the Proponent’s [potential partner] potential to form an alliance with the Owner’s personnel.” (ADIT 2011) The North American power industry has been using alli- ancing for at least a decade (Ray 2013) and has come to the same conclusion as the international public transportation agencies discussed previously. One insightful article on the the NOPs, its preferred collaborative alliance partners. The NOPs form teams that bid for the project and usually consist of contractors and consultants that, as a team, have the neces- sary expertise and size to take on the project. Once the agency selects the preferred alliance partners, they sign an interim project alliance agreement. In a pure alliance, the agency chooses a single team, and in a competitive alliance, the client chooses multiple teams that compete for the right to be the sole team to join the agency in the alliance. During the interim project alliance agreement phase the team(s) together with staff from the agency carries out pre- liminary design, development of the risk profile, prepares the project methodology, and prepares the proposed cost of the project, the TOC. The TOC is similar to the U.S. con- cept of a guaranteed maximum price and includes direct costs (e.g., investigations, permits, land purchase, design, construction, and commissioning), overheads, and profit margins. The financial drivers for the NOP teams are col- lected in Table 20 and show the differences between the alli- ance models. During the interim phase, decisions can be made on a “best-for-project” basis in both the pure and collaborative alliances, while in a competitive alliance, the short-listed teams compete to produce the “best” TOC proposal. This proposal has various qualification and experience param- eters as well as cost. In the final phase, there is no differ- ence between the competitive and pure alliances. However, because a collaborative alliance has more than one construc- tion team, the NOPs must compete for work packages as well as the allocated profit margins and final pain/gain sharing. Therefore, the pure alliance’s NOP financial drivers are the profit earned during construction and the pain/gain shar- ing at project termination. In a competitive alliance, compe- tition determines who gets to sign the final project alliance agreement. Hence the NOP’s financial drivers become the same as in the pure alliance. The collaborative alliance is dif- ferent in that the contractors compete to increase their share of work as a means to maximize their profit. This creates an incentive to continuously improve KRA ratings in areas of time, cost, quality, etc. Again, the pain/gain share drives best cost outcomes on each project work package and, because the pain/gain share is spread across all delivery teams, it also drives support between teams and sharing of ideas and advice TABLE 20 DRIVERS FOR THE NONOWNER PARTICIPANTS DURING THE INTERIM PROJECT ALLIANCE AGREEMENT PHASES IN THE THREE ALLIANCE FORMS Interim Project Alliance AgreemModel ent Project Alliance Agreement Pure Alliance best-for-project profit margin and gain sharing Competitive Alliance best value profit margin and gain sharing Collaborative Alliance best-for-project profit margin, gain sharing, and increasing forward workload Source: Scheepbouwer and Gransberg (2014).

44 and sufficient conditions.” The selection of the term “family resemblance” is somewhat unique when analyzing a major infrastructure construction project; however, the analogy is appropriate (Hauck et al. 2004). Thus, the question of how to form a collaborative “family” of design and construction professionals that each represents a different organization must be answered. The answer is easier to show than explain. Therefore, the next two sections contain cases where specific actions were taken to accomplish the same goal to form a highly collaborative team made up of members with both the ability and the will to make the collaboration a reality. The SCIRT Alliance in New Zealand illustrates a facilitated workshop approach to selecting alli- ance members. The National Museum of Australia project demonstrates the development of a work environment to keep the team after selection operating at a high level of productiv- ity and amicability. SCIRT Two-Day Alliance Team Selection Workshop The selection of alliance partners includes a two-day, off-site team evaluation workshop. Instead of evaluating individual personalities based on a resume, the whole proposed alliance (management) team is evaluated during scenarios that are played out during the workshop. Every alliance tender phase in Australia and New Zealand has this selection element in it. The scoring is part of the qualitative evaluation of the ten- ders and, although this process could be viewed as subjective, according to the NZTA in principle all qualitative or non-price elements in a tender evaluation are and they have not yet been challenged on the results (by the non-owner participants). The goals of this part of the evaluation are to see how the alliance operates as a team and under pressure, how they make deci- sions, how they interact with the owner, and how they interact with each other. An example of a scenario is as follows: All members of the management team of the alliance (except the owner staff) pretend to be on a construction site where there has been a fatality with one of the workers. The media heard about it and has called and wants to know what is going on and what you are going to do about it. In this scenario the proposed project director would be “on holiday” and is therefore taken out of the exercise. The rest of the people then have 20 minutes to tell the owner what the response would be on site, and how you would respond to the media. In some scenarios there is a full proposed alliance manage- ment team present; that is, including the members of the cli- ent organization and for some scenarios only the participating non-owner members are part of the exercise (possibly with key members removed as described previously). The workshop is led by facilitators specializing in alliancing and selection pro- cedures. The owner staff, among which include the tender eval- uation team and proposed alliance staff, observe the scenarios. After the two days the client team comes together and talks about the observations and tries to answer questions such as, benefits of alliancing in the power industry provides a quasi- public perception of alliance member selection: You have to be real careful about the partner you pick. It’s kind of like entering a marriage. You don’t want to do this very often. You just want to do it once and hopefully be done with it. The compa- nies need to have a shared value system and the cultures need to complement each other. That doesn’t mean it’s always rosy every day. Without a little bit of conflict, you’re probably not sharp- ening your saw and getting better . . . We challenge each other and we work to get better . . . You have to define the parameters around working toward that shared goal. You also have to pro- vide some kind of framework for resolving conflicts . . . If you’re going to be an alliance contractor, you have to see yourself as the brother-in-law sleeping on the couch. You better add value every day if you want to have a place to stay. (Ray 2013) This all leads to the conclusion that unlike other alternative project delivery methods where the emphasis on key person- nel qualifications focuses on their experience and credentials, the emphasis in the qualifications evaluation process for an alliance is focused on the personalities of the key personnel. While U.S. and Canadian agencies have dabbled in the “soft factors evaluation” by experimenting with problem-solving scenarios in the CMGC selection process (West et al. 2012), the inherent subjectivity of evaluating an individual’s person- ality to determine their ultimate qualification to join the alli- ance team may make this aspect very difficult to implement in North America. Alliance Team Building For an American agency, the notion that litigation is not an option in a construction contract is shocking at best and absurd at worst. However unattainable it may seem in the United States, that “one third of the total value of public sec- tor infrastructure projects delivered in Australia” are alliance contracts (Duffield et al. 2014) demonstrates that to effect real change there must be a will to change (Yeung et al. 2012). The conclusion reached in the previous paragraph underscores the need to approach complex projects with a keen focus on cre- ating a team consisting of representatives from each stake- holder who are willing to make the shift from the status quo to something better. Hauck et al. (2004) cited five factors that must be present to maximize collaboration among construction stakeholders: (1) high-performance teams, (2) optimization and performance measurement, (3) communication, (4) incentives and risk shar- ing, and (5) problem solving and decision making. Yeung et al. (2012) goes on to argue that successful relational contracts such as an alliance require the team members to possess the following five attributes: (1) commitment, (2) trust, (3) coop- eration and communication, (4) common goals and objective, and (5) win-win philosophy. That paper goes on to evaluate alliancing using the “Wittgenstein family-resemblance phi- losophy” that approaches complicated issues by defining them as “a network of overlapping similarities. This is dissimilar to the traditional definition whereby a concept is given necessary

45 ated. It included an “excellence allowance” that increased the workers’ hourly pay by as much as AU$1.75/hour for achieving key performance indicators for production, qual- ity, and safety. The agreement was attributed with increasing production by 30% (Hauck et al. 2004). To summarize this section’s take-away points, a success- ful alliance is one that is composed of faces with personali- ties not just positions with credentials. The SCIRT workshop demonstrated an effective way to “filter” the personalities of potential team members using scenario-based interactions. The Australian National Museum provided an approach for creating a work environment where those selected collabora- tive personalities can thrive. ALLIANCING PROCEDURES Although the literature review found a myriad of agency pro- cedures related to alliancing, it appears that two stood out as the most prominent and critical to understanding the prin- ciples of this new project delivery method; developing the “commercial framework” and developing the TOC. There- fore, this section will limit its discussion to these two topics. Commercial Framework The commercial framework is defined as the terms and con- ditions under which the project alliance agreement will oper- ate. These are directly connected to the statutes that govern procurement by the agency seeking to consummate a project alliance agreement. As such, there is no attempt to spell out specific language, but rather the emphasis is on the underly- ing principles based on English Common Law, which is the basis for both the Canadian and U.S. legal systems. More specific legal guidance is contained in chapter five of this synthesis. There are three factors that alliances address in their com- mercial framework. The first is to make the business case for the project as well as for delivering it using an alliance of the type proposed by the agency. All of the case example projects, except for Washington State, used the VfM method- ology for demonstrating that the proposed approach to form- ing an alliance and then delivering the project was the “best or preferred” alternative (ADIT 2011). The procedures are virtually the same as those used by U.S. DOTs during early stages of the project development process. The second factor is the required level of integration that must be achieved by the alliance itself (ADIT 2011). “The greater the degree of integration of the skills and dis- ciplines of its different members, the more likely it is that a collaborative approach is possible, and outstanding results achieved for the project sponsor (where applicable), project owner, and the team” (Austroads 2014). Therefore, the com- mercial framework must address organizational changes that do they perform well as a team and do they complement each other. According to the NZTA, sometimes there are strong but disparate personalities that just do not work well together. This whole exercise is repeated for every proposed alli- ance team that has reached this procurement phase. National Museum of Australia Alliancing Development Process Like SCIRT, the National Museum project took extraordinary steps to ensure that the members of the alliance were both competent and compatible. Each potential consultant or con- tractor was asked to prepare a statement of qualification that provided specific evidence that each member of the organiza- tion’s proposed team complied with the following 12 criteria: 1. Demonstrated ability to complete the full scope of works including contributing to building, structural, mechanical, and landscaping design. 2. Demonstrated ability to minimize project capital and oper- ating costs without sacrificing quality. (Value analysis and life-cycle costing.) 3. Demonstrated ability to achieve outstanding quality results. 4. Demonstrated ability to provide the necessary resources for the project and meet the project program. (Including resumes of key staff.) 5. Demonstrated ability to add value and bring innovation to the project. 6. Demonstrated ability to achieve outstanding safety perfor- mance. 7. Demonstrated ability to achieve outstanding workplace relations. 8. Successful public relations (PR) and industry recognition. 9. Demonstrated practical experience and philosophical approach in the areas of developing sustainability and envi- ronmental management. 10. Demonstrated understanding and affinity for operating as a member of an alliance. (Collaborative experience and views on risk/reward schemes.) 11. Substantial acceptance of the draft alliance document for the project including related codes of practice, proposals for support of local industry, and employment opportunities for Australian indigenous peoples. 12. Demonstrated commitment to exceed project objectives. (Hauck et al. 2004) Once selected the members of the alliance were required to collectively put their profitability at risk if these performance measures were not met. Because of this collective nature of the risk and reward incentives, no member of the alliance could succeed unless all members succeeded and the failure of one partner could directly threaten the profitability of all other alliance members. It is this joint, rather than just shared, risk and reward structure that distinguishes project alliances from other forms of contracting and partnering arrangements. (Hauck et al. 2004) The most effective example of developing a collaborative environment that penetrates to the grassroots of the project team was an incentive-based project labor agreement with the trade unions with which the craft workers were affili-

46 designs to choose another delivery model such as DB and solicit proposals or tender offers to complete the project. 2. In the Netherlands case, the TOC was developed with a ceiling of €140 million. This encourages the agency and its partners to work together until they agree on a conceptual design and project plan with a correspond- ing cost estimate that conforms to functional require- ments, the commercial objectives of the private-sector members, and the timeline desired by the agency. Once a number was agreed upon that fell below the cap, the parties were free to enter into a project alliance agreement. 3. In the competitive alliances used in Australia and New Zealand, prequalified project teams develop a project plan and respective TOC in much the same manner as a best-value DB proposal. Agency personnel are assigned to work with each competing team and pro- vide the same input that the agency would provide in a pure alliance when establishing the TOC. The TOC proposals are submitted and evaluated by the agency, which then selects the winning team to join the alli- ance. The award makes the TOC a contractual part of the project alliance agreement. 4. In a collaborative alliance (SCIRT) each individual project in the program of works receives a TOC. A conceptual TOC is produced during preconstruction services provided by the construction contractors dur- ing the design phase; however, in the end, the final TOC is set by the agency. It is important to note that the type of alliance determines the pay scheme for the alliance members. In the pure and collaborative alliances, a reimbursable payment scheme is used. This is done to achieve the principle that everyone wins or everyone loses financially. A reimbursable scheme covers may need to be made to achieve the required integration. For example, the SCIRT Alliance’s framework required each alli- ance member to second its personnel to the alliance, thereby changing their individual corporate identities to meld them into a single alliance identity (Scheepbouwer and Gransberg 2014). This involved making arrangements for collocation of all the personnel to SCIRT offices and issuing SCIRT-labeled hard hats, safety vests, shirts, etc., to reinforce the idea that they no longer worked directly for a contractor, a consultant, or an agency. This undertaking required an enormous admin- istrative effort by the human resources managers for each alliance member. The third factor is to align the commercial objectives of the alliance members. The owner’s typical commercial objec- tive is to complete the project on time at the lowest practical cost. The design consultant and the construction contractor have maximized the profit they earn on the job. Therefore, one way to align these two objectives is to create an incen- tive for cost savings as was done in the Port of Melbourne alliance case. This also takes a collaborative effort, so the agency must produce its procurement documents in a manner that allows their modification as new members are brought into the alli- ance with disparate commercial objectives. One common example found in most of the case example alliances was the mechanism for developing each member’s profit/loss share. Once again SCIRT furnishes a good example of how to accomplish this task on a very complex project. Once the col- laborative project alliance agreement was executed, the alli- ance’s audit arm conducted audits of each member’s books and determined their “usual margin,” the actual average profit they made on each completed project (Scheepbouwer and Gransberg 2014). These rates were then used as a starting point for filling that line item cost by negotiation. Developing the Target Outturn Cost The typical structure for a TOC is shown in Figure 16. It is important to note the difference between what are reimburs- able costs under the alliance’s open books accounting system and what is considered the NOPs’ fee. The literature and case studies yielded four methods to develop the TOC. Each is specific to both the agency involved and the type of alliance being undertaken. 1. In a pure alliance, the agency chooses a project team that then proceeds to develop a project plan of action for design and construction. Once the plan is solidi- fied, the team proceeds to develop the design to a point where a mutually agreeable TOC can be developed. After the TOC formation, the agency decides whether to continue or halt the project. If the decision is to halt the project, the agency is free to proceed with the FIGURE 16 Typical target outturn cost structure (ADIT 2011).

47 able and controlling them, and a framework for the effective assignment of specific and overall accountability for deliver- ing the project. It is a set of policies, principles, rules, and sup- porting practices put in place to run a project” (ADIT 2011). Figure 17 illustrates the structure of a typical alliance and serves as a generic example of how the hierarchy of gover- nance must be established. The Australians call this system a “joint management structure” and have detailed guidance on the roles and responsibilities of each of the individuals shown here. A brief synopsis of the major points about each group in Figure 17 is as follows. • Alliance Leadership Team (ALT): The primary rule of the ALT’s role is that all decisions assigned to the ALT must be made unanimously, in keeping with the “we all win or we all lose” philosophy. To work, each of the alliance members must assign a representative that is truly “the best person for the job.” This group is where the emphasis on key personalities becomes valuable and where the “no sue” rule is ultimately tested. The group’s scope of decision making must be well-defined and the agency normally reserves the right to make the project direct costs and the overhead costs for alliance members. The competitive alliance uses the winner’s pro- posed TOC as the alliance agreement’s contractual TOC. The competitive alliance was developed to counter criticism that the agency may not be getting good VfM. Therefore, it might be inferred that a reimbursable payment scheme is preferred unless the agency believes the need to demonstrate that it got VfM by implementing a competitive alliance. ALLIANCING PROGRAMS Once again there are a number of agency programs that attend to alliance contracting; however, the three described in this section appear to be the most unique. They are the alli- ance’s governance program, risk management program, and the painshare/gainshare scheme. Governance “Governance can be described as a process for directing and managing projects, a system for holding projects account- FIGURE 17 Typical alliance governance structure (ADIT 2011).

48 certain decisions unilaterally, especially those related to long-term operations and maintenance issues. • Alliance Manager (AM): This is the individual charged with leading the alliance and who is usually vested with the authority to make many routine decisions. The AM reports to the ALT and is typically a senior project manager from a NOP. The ALT may select a person from within its cur- rent base of member’s senior personnel or it may choose to hire a specialist with no affiliation to any of the mem- bers. The AM chairs the AMT. • Alliance Management Team (AMT): This group is the one that actually ensures that the project gets designed and built and is normally composed of members selected for their special expertise and experience. They usually come from the agency and NOP staff and are seconded to the alliance from their parent organizations. Risk Management “The most significant difference between traditional con- tracting methods and alliance contracting is that in alliancing, all project risk management and outcomes are collectively shared by the Participants” (ADIT 2011). The very core of an alliance is the equal sharing of all risks and rewards. Other project delivery methods advocate assigning risks to the party that can most effectively manage them and, as a result, that party is the sole recipient of any rewards or penalties based on how well it managed the risks. The Melbourne Channel Deepening case is a good example of an alliance with a comprehensive risk management plan that was jointly governed by the members of the alliance. This project was faced with a complex risk profile that included a bewildering array of potential external impacts that threatened the success of the project. The following is a brief review of the major risks and how the alliance dealt with each. • Environmental issues during dredging: An external independent environmental watchdog organization was set up to monitor the alliance’s compliance with more than 150 KPIs. • Negative public opinion: The environmental monitor regularly kept the public informed of the alliance’s record through reports and news releases. They also established a project stakeholder advisory committee to provide a conduit for information to the alliance and a platform for the alliance to respond to queries and concerns. • Disruption of ocean-going freight traffic: The alliance awarded DB contracts to specialty firms with specific expertise in the types of marine facility construction needed and coordinated the work with the Port of Mel- bourne to minimize disruptions. • Underwater construction safety: The alliance devel- oped a specific dive industry liaison group to coordinate the underwater construction and ensure safety standard compliance. Risk management programs are always directly related to the project-specific technical requirements as well as the environmental, social, and political context in which the project must be delivered. This leads to the conclusion that implementing an alliance will require many agencies to make a large shift in their traditional risk management programs away from risk shedding and risk allocating to true risk shar- ing to benefit from this project delivery method. Gainshare/Painshare Schemes The risk sharing philosophy discussed in the previous section is implemented in the specifics of the alliance’s gainshare/ painshare scheme. Typically, the basis of the scheme is found in the KRAs and the various KPIs used to measure perfor- mance of the outcomes. The scheme will include both cost and non-cost metrics, which are combined to calculate the shares of the each member’s gain or loss. Non-cost perfor- mance criteria are generally related to design and construc- tion quality, timely achievement of scheduled milestones, measures of traffic disruption, customer satisfaction, envi- ronmental compliance, safety, and other areas found in a typical large construction project. Figure 18 shows how two typical gainshare/painshare schemes are developed. The one on the left has no limita- tions on either the agency or the NOP’s gains or losses. The right-hand graph shows a more common scheme where the NOP’s losses are capped at the amount of fee (profit) they were accorded in the project alliance agreement (Queensland 2008; ADIT 2011). This scheme is Limb 3 of the typical 3-Limb pricing structure discussed in chapter two. Table 21 provides an example of how the left-hand scheme in Figure 18 is calculated for both the industry members and the agency. It contains a NOP incentive/disincentive mecha- nism for achieving the non-cost performance criteria in the KRAs. In this case it is ±AU$2 million. Appendix A contains an extract of the Australian National Alliance Contracting Guidelines and describes in detail the full set of options with example calculation available to public highway agencies in that country. FIGURE 18 Typical gainshare/painshare models (ADIT 2011).

49 CONCLUSIONS The following conclusions were developed in this chapter: • From Table 18—Projects that are good candidates for alliance contracting delivery are highly complex proj- ects worth AU$50 million (~U.S. $47 million) or more. They have high-risk profiles with a “potential for a sub- stantial change in project scope” (Queensland 2008) and therefore require flexibility to make decisions and change plans in an expeditious and agile manner. The TABLE 21 AUSTRALIAN DEPARTMENT OF INFRASTRUCTURE AND TREASURY MODEL FOR GAINSHARE/PAINSHARE EXAMPLE Cost Gainshare/Painshare 50:50, no caps Alliance Performance Score Reward/Penalty to NOP TOC Components Reimbursable Costs $88 M Achieve Stretch Target 100 $2 M NOP Aggregate Fee (Profit) $12 M Business as Usual 0 $ 0 TOC $100 M Poor –50 –$2 M Risk or Reward Calculations Model 1a) Cost and Non–Cost Performance Not Linked Scenarios 1 Very good cost and non-cost performance 2 Mixed—Very good cost and poor non-cost performance 3 Very poor cost and non-cost performance TOC $100 M $100 M $100 M AOC $90 M $90 M $125 M Under (overrun) to TOC $10 M $10 M –$25 M Non–Cost Performance Score 100 –50 –50 Cost Gainshare/Painshare Owner 50% +$5 M +$5 M –$12.5 M NOPs 50% +$5 M +$5 M –$12.5 M Non–Cost Reward/Penalty to NOP +$2M reward –$2 M penalty –$2 M penalty Total Gainshare/Painshare Owner +$5 M +$5 M –$12.5 M NOP +$7 M +$3 M –$14.5 M Source: ADIT (2011). risk profile is complicated by the large number of exter- nal stakeholders and often an aggressive schedule. • Alliance contracting procurement demands that a con- siderable amount of weight be placed on the person- alities of the key personnel, unlike other alternative project delivery methods where the emphasis on key personnel focuses on their experience and credentials. • Implementing alliance contracting will require North American agencies to shift their risk management pro- grams away from risk shedding and risk allocating to real risk sharing to benefit from this project delivery method.