Part 3: An alliance procurement strategy

3.1
In this Part, we assess whether the selection of an alliance as a procurement strategy in the context of the Canterbury earthquake recovery was appropriate. We also note some complexities brought about by the alliance model.

3.2
We looked at the rationale behind the three public entities' proposal to form an alliance to rebuild the horizontal infrastructure in Christchurch and how they selected their alliance partners.

3.3
We expected that the three public entities would have carefully considered which procurement approach was the most suitable for the situation and have based their recommendation on a compelling value proposition that supported the use of an alliance.

Summary of this Part

3.4
Selecting an alliance to carry out the horizontal infrastructure rebuild in Christchurch was, in our opinion, an appropriate choice for the circumstances that the three public entities were dealing with.

3.5
Because of the large scale of damage, uncertainty, and urgency of repairs, an alliance provided a useful approach for the three public entities to better manage the risks that would emerge from this situation than a more traditional style of contracting arrangement.

3.6
Although SCIRT exhibits many attributes typical of an alliance, having multiple owners and non-owners is unusual. Also, the extensive scale of the damage and a programme of works rather than a single project makes SCIRT more complex. We discuss how well the three public entities manage these risks in Part 5.

Selection of an alliance as the procurement method

The entities recognised a need to reassess their procurement model after the second major earthquake and used an established methodology to select an approach that was suitable for the circumstances.

3.7
A developing body of literature describes alliance contracting and discusses the circumstances in which it is an appropriate procurement method. The National Alliance Contracting Guidelines from the Australian Department of Infrastructure and Transport say that projects suitable for alliance contracting generally have one or more of the following characteristics:

• The project has risk that cannot be adequately defined or measured in a business case or before tendering.
• The cost of transferring risk is prohibitive.
• The project needs to be started as early as possible before the risks can be fully identified and/or project(s) scope can be finalised, and the owner is prepared to take the commercial risk of a suboptimal price outcome.
• The owner has superior knowledge, skills, and capacity to influence or participate in the development and delivery of the project.
• A collective approach to assessing and managing risk will produce a better outcome.

3.8
Our literature review also suggested that alliancing might be appropriate for reconstruction projects after a major earthquake because of the following factors:

• uncertainty about the availability and cost of people and other resources, which might limit competition;
• the large scale of the work programme;
• the large scale of uncertainty and complexity;
• the short time for rebuilding;
• the need to use local materials, labour, and plant;
• local industry familiarity with construction procurement and delivering construction projects;
• the need for a target-cost type of payment mechanism to allow for variation in the scope of work and promotion of innovation in its execution; and
• the generation of a co-operative culture because of the wider social incentives to work together for the benefit of the whole community.

3.9
CCC and NZTA considered that an alliance delivery model was a more suitable way of delivering the reinstatement works than scaling up the existing arrangements. (CERA was not involved because it had only recently been formed.) CCC and NZTA considered that the alliance model would foster a high degree of trust between parties and focus on high-performing expectations, because of the system of rewards and sanctions for achieving or missing mutually agreed targets.

3.10
It was thought that other possible models, such as "Design and Construct" or "Managed Contractor Model", would not deliver with the speed required, would have complex administrative layers, and would not effectively bring together organisations with differing objectives. CCC had experience and capability in asset management, and NZTA has had a number of successful experiences with alliancing. The situation matched the criteria for using an alliance and provided a useful approach for CCC and NZTA to manage the risks.

3.11
The alliance approach was also favoured because of the benefits that the approach usually delivers. These benefits include reduced overheads, streamlined approvals, increased private sector participation, increased agility to deal with an evolving scope of works, and achieving multiple objectives. An alliance would also provide incentives to reduce start-up times and finish ahead of schedule and encourage high performance in areas such as stakeholder relations, communication, maximising use of the local labour force and contractors, and increasing general workforce skill levels as a consequence.

Selection of alliance non-owner participants

The selection of non-owner participants was not a competitive process run specifically for SCIRT. CCC had selected the contractors through competitive tender a few months before SCIRT was formed. These contractors became the non-owner participants. Mechanisms are built into SCIRT's commercial framework to create competitive tension between non-owner participants, and there are opportunities for other contractors to carry out work.

3.12
Our literature review on alliance contracting showed that the process for selecting the non-owner participant(s) for an alliance is important. Alliance contracting envisages a competitive process that takes account of the requirements of alliancing and incorporates consideration of both price and non-price aspects.

3.13
In response to the September 2010 earthquake, CCC set up the Infrastructure Rebuild Management Office (IRMO) to project-manage the reinstatement of infrastructure. The IRMO comprised CCC staff. CCC selected the construction companies that would rebuild the damaged areas and entered into four design-build contract arrangements through a competitive tender process.

3.14
The situation changed on 22 February 2011, when another earthquake struck Christchurch much closer to the central city. The damage was more widespread, and CCC recognised that its arrangement was no longer suitable for the size and scale of the task. Appendix 1 describes the circumstances leading to the formation of SCIRT in more detail.

3.15
CCC and NZTA decided that the IRMO contractors would become the non-owner participants in SCIRT. They made this decision because of the contractors' recent selection for IRMO work, their local presence, and a judgement about the workable number of non-owner participants in SCIRT.

3.16
The number of non-owner participants was limited to five, because CCC and NZTA considered this to be the maximum number for a well-functioning alliance. It was noted that other contractors might wish to be involved as head contractors, but five made sense in terms of the scale of their local presence and workable alliance function.

3.17
Mechanisms are built into SCIRT's commercial framework to create competitive tension between non-owner participants, encourage high levels of performance, and constrain cost inflation. SCIRT is unique in that it is designed to have both competitive and collaborative mechanisms operating concurrently. We evaluate the effectiveness of these tensions in Part 4.

3.18
The selection of non-owner participants was not a competitive process carried out specifically for SCIRT, but we have no means of assessing whether a more competitive process would have yielded a better result. The principles of a competitive process were met when lead contractors were selected for IRMO during the previous year. In our view, it was a practical decision to carry the lead contractors over from IRMO into SCIRT as non-owner participants. We also note that the Alliance Agreement requires non-owner participants to subcontract a minimum of 40% of the work, which gives opportunities to other contractors.

Complexities of SCIRT as an alliance

SCIRT is more complex than a usual alliance because it involves multiple owners and non-owners, an unprecedented scale of damage, a programme of projects, and a team formed entirely from contractors and secondees. The three public entities and SCIRT have taken steps to address this complexity, but there are ongoing risks that they will have to manage during the rebuild.

3.19
Although SCIRT exhibits many of the attributes typical of an alliance, it is more complex than most. The presence of multiple owners is an unusual feature and immediately makes planning and decision-making more complicated. The three public entities have responded by creating the CGG, which provides a forum for them to discuss issues pertinent to the asset owners and funders.

3.20
The extent of damage to buildings and infrastructure in Christchurch is unprecedented in New Zealand. Ultimately, about 1500 buildings will be demolished in the CBD, making it the country's largest construction site. More than half of the urban roads were damaged, and there is damage to almost 800km of water reticulation infrastructure. Most of this damage is to the wastewater network. A typical NZTA alliance, such as construction of the Waterview Connection in Auckland, consists of only one project. SCIRT consists of hundreds of projects in a programme of works. SCIRT's operational systems have been customised to deal with this complexity, although we were told that getting these systems up and running does not happen all at once. In our view, there are ongoing risks that will have to be managed during the rebuild. We discuss how well the three public entities are managing these risks in Part 5.

3.21
One of the main challenges for SCIRT is to bring together a mixture of staff from different organisations to work together as a team and deliver for the people of Christchurch. As performance is important to SCIRT, the SCIRT Management Team has prepared a Peak Performance Plan that is renewed annually. Its purpose is to provide SCIRT with a strategic map for building and sustaining outstanding performance.

3.22
The Peak Performance Plan is monitored through engagement surveys and exit interviews. It was externally reviewed in 2012.⁶ The review concluded that the Plan was having a positive effect and also remarked:

SCIRT is best described as an organisation that despite operating in a complex and uncertain environment has a clear sense of purpose, an outcome focus and a team of aligned and committed members.

In our view SCIRT has made extraordinary progress towards its goals over a very short timeframe; undoubtedly the Board and leadership teams' focus on both creating and expecting a culture of high performance has been an integral part of SCIRT's success.

One of the key strategic mechanisms that SCIRT has used to achieve these results has been its Peak Performance Framework … It is a best practice example of intentionally designing key organisational structures and processes to develop a high performance culture.

3.23
Supported by the KRA framework, SCIRT aims to set high standards and drive improvement in safety, environment, quality, community, and stakeholder relationships. SCIRT has a focus on positively affecting the social well-being of the city and the people of Christchurch.

6: External review carried out by the University of Canterbury, NZTA, and Idea Creation Limited.