Selection was a three-stage process:
• EOI phase - this four week phase served to register proponents' interest.
• SIA (Statement of Interest and Ability) phase - in this two month phase, those parties who had completed the EOI stage were invited to provide a response addressing their experience and track record, as well as technical and management skills. Responses were assessed on non-price criteria. The proponents' affinity for working collaboratively was assessed within the above criteria.
• RFP Phase - Two proponents selected from the SIA phase were invited to participate in this phase - the TOC Development Phase - culminating in the submission (by each proponent) of a Project Proposal. The RFP was issued in March 2007, with proposals due by July 2007, four months later.
The non-cost attributes (evaluation criteria) used to assess the Project Proposals were:
• Relevant Experience
• Track Record
• Technical Skills
• Resources
• Management Skills
• Methodology
In differentiating the two proponents participating in the TOC Development Phase, the Evaluation Panel used the information provided in the Project Proposals to estimate a "Net TOC". This figure, used for evaluation and comparative purposes only, was calculated by adjusting each of the TOCs (up or down) to reflect the following specific factors:
• A "Supplier Quality Premium" assessed against parameters stated in the RFP. The process - as currently used - is described in the NZTA's procurement manual (available at <www.nzta.govt.nz>) and provides a mechanism for considering non-cost criteria.
• A Target Completion Date earlier than the "not to exceed" date specified by NZTA, with the adjustment calculated using an "adjustment amount per week" advised in the RFP.
• The difference in 'value' (whole of life costs, benefits and risks) between the quality of product offered by the proponents. For example, a superior urban design solution could have earned an adjustment of up to NZ$200,000.
• The difference between the proposals in relation to the risk-adjusted cost associated with performance against KRAs (other than Time, which was covered by the adjustment for Target Completion Date)
• Risks transferred to NZTA, where different from the risk allocations outlined in the draft PAA.
The proponent submitting the proposal assessed as having the lowest "Net TOC" would become the Preferred Proponent.
Through the process described, all elements of the project were developed in the competitive environment except the risks associated with resource consents (eg from Local Councils in respect of land use). NZTA progressed the resource consents independently and advised the outcomes to the proponents, with the understanding that in the event that any differences between the actual conditions and the expected conditions had a significant cost impact, NZTA would consider issuing a variation to amend the TOC.
In terms of whether the dual TOC proved cost-effective, NZTA is convinced that it did. During the TOC development phase, the two shortlisted proponents identified, explored and assessed a range of potential solutions, in the full knowledge that they were effectively in a competition to generate a solution with the lowest "Net TOC".
The primary costs to NZTA - for the dual TOC process compared to a single TOC process - are discussed below (generally without actual figures or percentages, in view of confidentiality and of the material variability that is to be expected between projects):
(i) the amounts paid to the two shortlisted proponents compared to the amount that would have been paid to a single proponent; NZTA paid both proponents a sum intended to cover 50% of bid costs.
(ii) offsetting the cost for two establishment audits rather than one, the audit scope was smaller because there was no need to validate margins (which were bid competitively by the proponents).
(iii) whereas an independent estimator is generally considered essential for a single TOC process (to ensure the estimates are founded on realistic costs and rates), such a role is unnecessary in a dual TOC process due to the competitive pressure. However, such savings were somewhat negated by the need to engage an independent estimator to work with NZTA in the "target adjustment" process used to compare the Project Proposals.
(iv) legal costs were marginally higher, due to the need for refining two IPAAs and two draft PAAs, but not by a large amount (noting that the interests of probity, the two IPAA documents would be unlikely to differ materially).
(v) Owner Team personnel time/costs were significantly higher in comparison with what would be expected under a single TOC process, though not double. However, the difficulty in making available the appropriate skillsets for the duration required should not be under-estimated.
(vi) the costs for probity advisers were higher than for a single TOC process, since the effective duration of the selection process (the primary period of demand for probity resources) was far longer than for a single TOC process. (However, this additional cost is minor in the totality of the project establishment costs.)
Overall, improvements and cost reductions incorporated in the winning proposal, when compared to NZTA's Specimen Design (the reference design, albeit incomplete), indicated increased value for money which more than compensated for the additional costs of the dual TOC process. For example, whereas the reference design contemplated a significant cost for landfill charges, the winning proponent devised a solution which extensively recycled waste from the existing highway. Similarly, innovative pile-cutting equipment developed for the project achieved productivity rates that had not been anticipated in the reference design.