South African National Road Agency Ltd v The Toll Collect Consortium and Another (796/2012) [2013] ZASCA 102; [2013] 4 All SA 393 (SCA); 2013 (6) SA 356 (SCA) (12 September 2013)

REPORTABLE

THE SUPREME COURT OF APPEAL OF SOUTH AFRICA

JUDGMENT

Case no: 796/2012

In the matter between:

SOUTH AFRICAN NATIONAL ROADS

AGENCY LIMITED ...........................................................................Appellant

and

THE TOLL COLLECT CONSORTIUM .............................First Respondent

TOLCON LEHUMO (PTY) LTD ......................................Second Respondent

Neutral citation: SANRAL v The Toll Collect Consortium(796/2012)[2013] ZASCA 102 (12 September 2013)

Coram: NUGENT, BOSIELO, WALLIS and PETSE JJA and SWAIN AJA.

Heard: 30 AUGUST 2013

Delivered: 12SEPTEMBER 2013

Summary: Tender – review of decision to award – transparency and objective standards – what constitutes –

no need to disclose basis for evaluating tenders.

ORDER

On appeal from: KwaZulu Natal High Court, Durban (VahedJ sitting as court of first instance) it is ordered that:

1. The appeal is upheld with costs, such costs to include the costs consequent upon the employment of two counsel.

2. The order of the court below is set aside and replaced with the following order:

‘The application is dismissed with costs, such costs to include those consequent upon the employment of two counsel.

WALLIS JA (NUGENT, BOSIELO and PETSE JJA and SWAIN AJA concurring)

[1] The appellant, SANRAL, which has responsibility for South Africa’s national roads, invited tenders for the operation of the N2 South Coast Toll Plaza. Four tenderers submitted tenders, and the contract was awarded to TolconLehumo (Pty) Ltd, the second respondent. The first respondent, a consortium of three companies operating under the name of The Toll Collect Consortium (the Consortium), was dissatisfied at the fact that its tender was unsuccessful and brought review proceedings in the KwaZulu Natal High

Court, Durban, challenging the award of the tender. The review succeeded.Vahed J set aside the award and ordered SANRAL to reconsider the tenders in the light of his judgment, subject to the qualification that no-one involved in the original assessment of the tenders should be involved in the fresh evaluation. This appeal is with his leave.

[2] The tender documents provided that the assessment of tenders would take place in two stages. First they would be assessed for quality and given a score out of 100. Failure to receive 75 or more points would result in the automatic disqualification of the

tender irrespective of price or other considerations such as Black Economic Empowerment (BEE). Thereafter the remaining tenders would be assessed based on price, contribution to BEE and other matters. The Consortium’s tender was disqualified at the first hurdle in that it received only 64.49 points on the quality assessment. It accepted that unless it could successfully challenge this assessment its tender was properly disqualified and the review could not succeed. The primary issue in considering the appeal must accordingly be the complaint in regard to the quality assessment. However, it is first necessary to deal briefly with two preliminary points raised by SANRAL

[3] The first point was that the Consortium lacked locus standito institute these proceedings under that name. There is no merit in this. If it was permissible, as it undoubtedly was, for the Consortium to tender for this contract under that name and title, as representing the three companies making up the Consortium, it must likewise have been permissible for the Consortium, under that name and title and still representing the same three companies, to challenge by judicial review the award of the tender. The second point related to the jurisdiction of the court below. As the contract flowing from the tender fell to be performed within its jurisdiction it clearly had jurisdiction to deal with a challenge to the award of that contract by way of judicial review. That point was also bad. I turn then to consider the issue in relation to the evaluation of quality under the tender.

[4] The tender documents dealt with the quality assessment in the following way. Item F3.11.3 in the tender data said:

'The score for quality will be based on the evaluation of the management proposals submitted by the Tenderer in terms of returnable

schedule Form D4.'

Item F3.11.3 went on to set out three broad quality criteria and the maximum score in respect of each criterion, those being 45 points for toll operations and maintenance; 50 points for toll systems and maintenance and 5 points for electrical and mechanical maintenance. No further breakdown was given of how the scores would be assessed by the independent adjudicator appointed by SANRAL for that purpose.

[5] Form D4, attached to the tender, set out in some detail a number of items that had to form part of a Management Proposal by prospective tenderers.Tenderers were required to provide a clear statement in respect of ‘at least the following information’ using the headings and sub-headings listed in a table forming part of the form. The table was divided into three main parts headed ‘toll operations’, ‘toll system’ and ‘electrical and mechanical system’ respectively. Plainly those covered the three heads under which the quality score was to be assessed. Each head was then broken down into a number of

sub-headings. Toll operations dealt with organisational structure, financial management, traffic management, risk management, quality

assurance, environmental management, safety and security and ‘other’. Toll system covered organisational structure, financial standing, risk management, quality assurance, environmental management, past performance, a detailed toll system roll out programme and ‘other’. Electrical and mechanical systems repeated five of the items appearing under the other heads together with ‘other’. There was accordingly a substantial overlapamong the three heads. This does not appear to have disconcerted the Consortium as its tender dealt with these items exhaustively in 67 pages, together with several hundred pages of annexures.There is no evidence to suggest that the Consortium queried these items or sought any additional explanation as to the manner in which they would be taken into account in adjudicating quality.

[6] The tender evaluation was first undertaken by an independent firm called Tolplan Operations (Pty) Ltd and then a formal tender evaluation was undertaken under the direction of SANRAL’s regional manager in Pietermaritzburg. Each of the three components that were to be scored was further broken down and the available points distributed among the various sub-headings. Thus ‘toll operations’ was separated into organisational structure,which was allocated 20 points, and proposed operations management, to which the remaining 25 points were allocated. Under organisational structure 5 points each were allocated for the proposed route organogram and the quality of key personnel and the remaining 10 points were given for staff salaries. The first four sub-items under proposed operations management were each allocated 5 points and the last two 2.5 points.

[7] A similar exercise was undertaken in respect of toll systems. Each of the sub-items identified in D4 under this head was allocated marks totalling 20 in all. The balance of 30 marks was reserved for ‘toll system technical analysis’. This latter item was treated as the ‘most critical management and control tool’ for the project. The Consortium did not raise any objection to this item or to the score that it was allocated under it. The complaints it makes in respect of its score for toll systems repeat its complaints in respect of three corresponding items under ‘toll operations’ and they can conveniently be dealt with

together with those items. Under the last heading of maintenance of technical and mechanical systems the Consortium lost points in relation to its technical manager, technician and non-use of a specialist sub-contractor.

[8] In relation to this scoring exercise Mr Ibrahim, who deposed to the affidavits on its behalf, said:

‘… I shall not perform an analysis of the points scored by the successful tenderer (ie the Second Respondent).The issue is not whether the Applicant ought to have scored more points than the Second Respondent for quality but rather whether the Applicant should have met the threshold of 75% for quality. It may, however, in some instances be necessary to draw a comparison between the Applicant’s points and the Second Respondent’s points to indicate that the points awarded to the Applicant were erroneous.’

Before dealing with that contention it is necessary to address what became in argument the primary issue raised on behalf of the Consortium. It is an issue that logically precedes any criticism of the scoring, because it claims that the process was deficient from the outset in consequence of a non-disclosure to tenderers.

[9] The foundation of the contention was the fact that the more detailed breakdown in the scoring system for quality described above in paras 6 and 7was not disclosed to tenderers. The heads of argument on behalf of the Consortium expressed the complaint in the following way. The absence of the breakdown in scoringwas said to result in the tender adjudication processbeing neither transparent nor objective. It was submitted that the obligation to disclose this information flowed from SANRAL’s Supply Chain Management Policy and Procedure Manual.Item 1.9.3.3of the Manual said,in regard to assessments of functionality (quality) such as this one, that when inviting bids an institution must indicate the evaluation criteria for measuring functionality; the weight of each criterion and the applicable values. Some reliance was also placed on the requirement of the Preferential Procurement Policy Framework Act (the PPPFA).1The submission in the heads of argument was that the three broad headings did not comply with these requirements, as ‘none of those headings would themselves be measured at all’. On that footing it was submitted that the tenderers should have been furnished with the weights and values of each sub-item in advance of submitting their tenders.

[10] SANRAL contended that the point was not open to the Consortium because it was not foreshadowed or developed in the affidavits delivered on its behalf. Furthermorethere was no indication that any prejudice flowed from the non-disclosure of this detail. It rejected the contention put forward in the heads of argument on behalf of the Consortiumthat this afforded an advantage to the existing operators who tendered because they ‘would have a better understanding of what criteria were regarded as important to SANRAL, and the relative importance of them.’ Lastly it contended that the invitation to tender indicated with sufficient

clarity how the overall scoring for quality would be undertaken and gave clear instructions as to the areas of concern to SANRAL so that all tenderers were aware of the information they needed to put forward in support of their tenders.

[11] In support of this final point SANRAL relied on the passage inMinister of Environmental Affairs and Tourism & another v Scenematic Fourteen (Pty) Ltd,2where Scott JA said:

‘A further point made by the respondent was that the applicants for fishing rights ought to have been told in advance of the procedure to be adopted, involving as it did the streaming of the applications into two groups and the use of a scoring system applied to predetermined criteria. It was argued that the failure on the part of the DDG properly to advise applicants rendered the allocation process procedurally unfair. Section 3(2)(a) of PAJA expressly provides that what is procedurally fair depends on the circumstances of each case. In the present case the applicants for fishing rights were required to complete a detailed application form which indicated precisely what information was required. It was accompanied by instructions on how to complete the form and guidelines setting out in broad terms the considerations which the decision-maker regarded as material for the purpose of making the allocations. An applicant would therefore have been fully aware of the information that was required and on which the allocations were to be made. In these circumstances, the decision-maker, in my view, was not required to explain in advance exactly how the applications would be processed. As Baxter Administrative Law at 548 puts it:

“The administration cannot be expected to share with the individual every phase of its final decision-making process.”’

[12] The Consortium’s contentions found favour with the court below. It held that it was incumbent on SANRAL to set out its benchmarks ‘up front’ so that a tenderer would know how to achieve the pre-determined scoring. The learned judge summarised his conclusion as follows:

‘It seems to me therefore, that if the first respondent wanted to assess and score quality and functionality on that basis, and more especially if quality and functionality was to serve a gatekeeper function, objectivity, rationality and functionality demanded more clarity on how prospective tenderers could get through the gate.’

For that reason the award of the tender was set aside and SANRAL was ordered to reassess the tenders in the light of the judgment but without using the same individuals to make the assessment.

[13] Two problems emerged with this order in the course of argument. The lesser one was that there was no basis for disqualifying the persons who had initially evaluated the tender from re-evaluating it. They were not accused of bias or mala fides or any other conduct that would warrant their disqualification. Accordingly that order should not have been made. A court should not lightly

disqualify officials and others who have the responsibility for administrative tasks and acts, such as the evaluation of tenders,

from performing their duties. It emerged from counsel’s submissions that the order had been sought in the light of an alleged

failure by the relevant officials to co-operate in providing information, butthat problem had been resolved before the case was

argued. Continuing to seek their disqualification was accordingly an error.

[14] The major problem was that the order did not address the difficulty the court identified in regard to the tender, namely the failure to inform tenderers in advance of the detailed breakdown of the possible scores under each of the headings used in the evaluation of the tenders. If there was a problem with the failure to afford tenderers adequate insight into the evaluation process it would not be resolved by remitting the tenders, formulated without the benefit of that information, for re-evaluation. Only setting aside the tender process and requiring SANRAL to undertake it afresh on the basis of adequate information could resolve it. But that relief was not asked for.

[15] Thishighlighted the fact that the point on which the review succeeded was not raised in the application papers, notwithstanding the fact that the applicant filed both a founding affidavit and a supplementary founding affidavit after receiving the record.It is so, that here and there complaints were made that the scoring process was subjective and that SANRAL had not stipulated any objective criteria for the evaluation of quality, but these complaints did not raise squarely the issue of the adequacy of the information afforded to tenderers about the more detailed scoring system to be used in the quality evaluation process. Had it been

raised then some attention might have been paid to the appropriateness of the relief being sought.

[16] In addition had the point been raised it would have compelled the Consortium to indicate in what way it was prejudiced by this failure. Counsel candidly accepted that there was not a word in the affidavits explaining how the disclosure of these details would have affected the Consortium’s tender. We were asked to infer prejudice from a hypothetical example drawn from the allocation of points in respect of the experience of some key personnel. However, in the absence of evidence that the Consortium would have looked to employ different key people had it been aware of this allocation, the point remained hypothetical. It did not establish that the Consortium suffered prejudice or that the tender process was flawed in a manner calculated to cause it prejudice. Absent special circumstances, prejudice is necessary in order for the applicant to demonstrate their interest in remedying by way of judicial review a fault in a tender process. Otherwise the court is being asked to deal with an academic issue.3 As this court said in Allpay:4‘It would be gravely prejudicial to the public interest if the law was to invalidate public contracts for inconsequential irregularities.’

[17] Leaving these problems aside and turning to consider the merits of the argument, the premise underpinning it is flawed. That premise amounted to this. In order for the tender process to be fair to all tenderers, SANRAL needed to disclose in advance, in the tender documents, full details of every element of the tender evaluation process that would be undertaken once tenders were submitted, including details of the breakdown in the allocation of the points for quality under the three heads set out in the tender documents.5 The submission was that this was necessary in order for the process to be transparent and for the evaluation of the competing tenders to be objective. Neither contention can be sustained.

[18] Transparency in a tender process requires that the tender take place in an environment where it is subject to public scrutiny. In other words the tender must be advertised publicly and its terms be available for public inspection. Those terms must set out clearly what must be submitted by those competing for the award of the contract. The adjudication of the tender must take place in an impartial manner and the results made publicly available. If there is a challenge to the outcome of the tender there must be a record that discloses how the process of adjudication was conducted. In that way the tender process is transparent and the public can see that it was conducted fairly.6 When the Constitution,in s 217, requires that the procurement of goods and services by organs of state shall be transparent, its purpose is to ensure that the tender process is not abused to favour those who have influence within the institutions of the state or those whose interests the relevant officials and office bearers in organs of state wish to advance. It requires that public procurement take place in public view and not by way of back door deals, the peddling of influence or other forms of corruption. But, once a tender is issued and evaluated and a contract awarded in an open and public fashion, that discharges the constitutional requirement of transparency. It is not there to be used by a disappointed tenderer to find some ground for reversing the outcome or commencing the process anew, by claiming that there should have been greater disclosure of the methodology to be adopted in evaluating the tenders.

[19] An attempt was made to draw a parallel between a student writing an examination and needing to know how many marks are to be allocated to each answer and the position of the tenderer needing to know where to concentrate in furnishing the information required by the party issuing the tender. But thetwo are not parallel situations. An examinee is subject to the pressures of time and must

complete a prescribed number of questions within a limited period. Part of the skill in writing exams is therefore to know what must be included and what can safely be omitted in answering a particular question. Tenders are not prepared under time constraints and the tenderer is free to put every possible scrap of potentially useful information into the tender. The tender can be as detailed as the tenderer wishes to make it. If there is a parallel to the academic situation it is rather with an essay or dissertation where the subject matter is identified in advance and the broad parameters outlined, but the examinee is free to respond as they see fit and in as much or as little detail as they choose.

[20] As to objectivity, which is an aspect of the constitutional requirement that the public procurement process be fair, it requires that the evaluation of the tender be undertaken by means that are explicable and clear and by standards that do not permit individual bias and preference to intrude. It does not, and cannot, mean that in every case the process is purely mechanical. There will be tenders where the process is relatively mechanical, for example, where the price tendered is the only relevant factor and the competing prices are capable of ready comparison. The application of the formula for adjudicating preferences under the PPPFA may provide

another example. However, the evaluation of many tenders is a complex process involving the consideration and weighing of a number of diverse factors. The assessment of the relative importance of these requires skill, expertise and the exercise of judgment on the part of the person or body undertaking the evaluation. That cannot be a mechanical process. The evaluator must decide how to weigh each factor and determine its significance in arriving at an appropriate decision.7Where that occurs it does not mean that the evaluation is not objective. Provided the evaluator can identify the relevant criteria by which the evaluation was undertaken and the judgment that was made on the relative importance and weight attached to each, the process is objective and the procurement process is fair.

[21] Where the evaluation of a tender requires the weighing of disparate factors it will frequently be convenient for the evaluator to allocate scores or points to the different factors in accordance with the weight that the evaluator attaches to those factors. But the adoption of such a system, without its being disclosed to tenderers in advance, does not mean that the tender processis not objective. If anything the adoption of the scoring system enhances the objectivity of the process, because, in the event of a challenge to the award of the tender, the basis upon which the evaluation was undertaken emerges clearly.

[22] The prior disclosure of any such points system – assuming that it was adopted in advance of the evaluation process and

not in the course of that process – is not ordinarily required, provided that the basic criteria upon which tenders will be evaluated are disclosed.8That is what this court held in Scenematicin the passage quoted above in para 11. Provided the tender documents make clear to participants what is required from them their task is to submit that information for evaluation. If they do not do so,or the information is inadequate when scrutinised,they run the risk that on that aspect their tender will fare less well. This is what happened in this case with two of the experienced tenderers. One of them gave no details in respect of its toll plaza manager and was awarded no points under this head. The other indicated that its operations manager was relatively new to the role and received fewer points as a result. Disclosure of any such refined process of scoring in relation to a tender evaluation process will only be required if its non-disclosure would mislead tenderers or leave them in the dark as to the information they should provide in order to satisfy the requirements of the tender. There is no evidence that this is the case here.

[23] The position in this case is that all potential tenderers knew that they would have to achieve 75 points for quality if their tenders were not to be excluded at the first stage of evaluation. They knew what matters they had to address in their tenders. The Consortium experienced no difficulties in that regard because it did not seek clarification from SANRAL or the independent evaluator in regard to the information that they needed to provide. It knew that a high priority would be placed on experience because

clause F2.1 of the tender documents said that only tenders from parties experienced in toll operations and maintenance would be

considered. As it lacked such experience, at least in relation to the operation of tolls for SANRAL, it needed to present the strongest

possible case in that regard. None of that would have been affected by knowing in advance that the 45 points for toll operations would be divided as to 20 for their organisational structure and 25 for their proposed operations management. Nor would the disclosure of the breakdown of points in other categories or within those categories have affected matters or altered the terms of their tender in any way.

[24] It is no surprise therefore that the Consortium did not complain in its affidavits about not knowing the detailed breakdown in the points allocation, within the three categories identified in the tender documents. This complaint, raised in argument and unsupported by any factual basis, is without merit and should have been rejected.

[25] After the evaluation report had been made available the Consortium delivered a supplementary founding affidavit. In summarising its contentions the deponent said that the adjudication report was incorrect in the manner in which it evaluated the Consortium’s tender for quality and that if its tender had been properly evaluated for quality it would have achieved the 75 points threshold and should not have been disqualified. The deponent proceeded to highlight various elements of the quality evaluation and criticise the points allocated to the Consortium in respect of these items. Although disclaiming an intention to engage in an analysis of

the points or to suggest that the Consortium should have scored better than the ultimately successful tenderer the deponent said that he would draw some comparisons ‘to indicate that the points awarded to the [Consortium] were erroneous’.

[26] Notwithstanding this disclaimer, this was in truth an invitation to the court to examine the evaluation exercise and determine a revised score for the Consortium that would enable it to cross the quality threshold. Explanations contained in the report, to

the effect that the Consortium lacked relevant toll operation experience or had not provided information to indicate that they were capable of managing and operating this type of contract, were roundly criticised on two grounds. First it was claimed that the project was a small one with limited risk and ideal for the purpose of introducing a new operator, even though they lacked experience in toll operations in comparison with their rivals. Second the background and experience of the three key individuals were highlighted and it was submitted that ‘it is clear that the quality of the [Consortium’s] personnelexceed that required for the management of the Contract’ and accordingly it should have been allocated a far greater number of points under this head. Had that occurred the 75 points threshold would have been surpassed.

[27] The invitation to re-score the Consortium’s tender for quality must be declined. Once again it must be stressed that this is not the function of a court. The task of evaluating and awarding these tenders rested in the hands of SANRAL, not the court,

and its decision must be respected, provided it was arrived at in accordance with the constitutional requirements applicable to public procurement as set out in s 217 of the Constitution, any applicable legislation and the terms of the tender. The court could only interfere if the process was infected with illegality. The courtwill not hesitate to interfere with the award of a tender where there is impropriety or corruption. However, where the complaints merely go to the result of the evaluation of the tender a court will be reluctant to intervene and substitute its judgment for that of the evaluator. It may not interfere merely because the tender could have been clearer or more explicit. Nor will it interfere because it disagrees with the assessment of the evaluator as to the relative importance of different factors and the weight to be attached to them. The court is only concerned with the legality of the tender process and not with its outcome.

[28] When one analyses the contentions on behalf of the Consortium they amount to nothing more than a different view of the merits of its tender to that taken by the independent evaluator and the evaluation committee. In regard to the contention that this was a relatively small project SANRAL’s approach was that it could not afford to have a toll road contractor that was unable to meet the standards of management and operations that it required. It accordingly adopted the stance that the first and foremost

consideration in awarding the tender was its assessment of the capability and capacity of the tenderer to carry out the tender

successfully. That was a different priority to the one identified by the Consortium, but it cannot be said to be illegitimate.In regard to the skills and experience of its key personnel one would expect the Consortium to sing their praises. However, the evaluators took a less sanguine view of them. That was a matter of judgment and it was one that was open to the evaluators. The court cannot interfere with that decision. Overall SANRAL may have adopted an overly cautious attitude, as the Consortium contends, but that is not a ground for review of its decision.

[29] It follows that the application for review should not have succeeded. Accordingly the following order is made:

1 The appeal succeeds with costs, such costs to include those consequent upon the employment of two counsel.

2 The order of the court below is set aside and replaced with the following order:

‘The application is dismissed with costs, such costs to include those consequent upon the employment of two counsel.’

M J D WALLIS

JUSTICE OF APPEAL

Appearances

For appellant: B E LEECH SC (with him A J BOULLE)

Instructed by: Werksmans Inc, Sandton

Symington & De Kok, Bloemfontein

For respondent: P J OLSEN SC (with him V VOORMOLEN)

Instructed by:Cox Yeats Attorneys, Durban

McIntyre & Van der Post, Bloemfontein.

1Act 5 of 2000. SANRAL was at the time exempted from complying with that statute, but its procurement policy indicated that it undertook procurement in accordance with the spirit of that Act. Accepting this, it is unnecessary to consider the extent to which the Act was in law binding upon SANRAL.

2Minister of Environmental Affairs and Tourism & another v Scenematic Fourteen (Pty) Ltd[2005] ZASCA 11; 2005 (6) SA 182 (SCA) para 18.

3Rajah & Rajah (Pty) Ltd & others v Ventersdorp Municipality & others 1961 (4) SA 402 (A) at 407H-408B.

4Allpay Consolidated Investment Holdings (Pty) Ltd & others v Chief Executive Officer, South African Social Security Agency &others 2013 (4) SA 557 (SCA)para 21.

5For the purposes of this judgment I accept that the detailed breakdown of points in respect of the different sub-headings was known to SANRAL when the tender documents were issued, although it is unclear from the evidence whether this was in fact the case.

6The UNCITRAL Model Law on the Procurement of Goods, Construction and Services (1994) was adopted by the Commission (the United Nations Commission on International Trade Law) in 1994. It has been influential in the establishment of public procurement regimes in many countries and it requires transparency in public procurement. The principle of transparency is satisfied if the law relating to public procurement is available; if tenders are publicly advertised; if the qualification requirements for tenderers, the subject matter of the procurement and the criteria for evaluation are specified; if information on any modifications of the tender is available; if details of the tender awards are published; if there is a right to be present when tenders are opened; if there is a record of the tender process and reasons are given for the acceptance and rejection of tenders. S P de la HarpePublic Procurement Law: A Comparative Analysis (unpublished LLD thesis in the University of South Africa, 2009) at 96-114; 424-438 and 583.

7MEC for Environmental Affairs and Development Planning v Clairison’s CC [2013] ZASCA 82 paras 20-22.

8The criteria and procedures that must be disclosed to tenderers under the UNCITRAL Model Law (Art 6(1)) relate to matters such as technical qualifications, legal capacity, solvency, good standing with the revenue authorities, absence of criminal records and the like. De la Harpe, supra, 102 fn 36. They do not extend to disclosure of a scoring system used in the assessment of these matters.