South African Petroleum Industry Association v Fuel Retailers' Association (28818/2014) [2023] ZAGPJHC 1301 (13 November 2023)

REPUBLIC OF SOUTH AFRICA

IN THE HIGH COURT OF SOUTH AFRICA

GAUTENG DIVISION, JOHANNESBURG

Case No: 28818/2014

NOT REPORTABLE

NOT OF INTEREST TO OTHER JUDGES

In the application for leave to appeal:

THE SOUTH AFRICAN PETROLEUM INDUSTRY ASSOCIATION Applicant And THE FUEL RETAILERS’ ASSOCIATION Respondent In re: THE FUEL RETAILERS’ ASSOCIATION Applicant And THE MINISTER OF ENERGY First Respondent THE CONTROLLER OF PETROLEUM PRODUCTS Second Respondent THE SOUTH AFRICAN PETROLEUM INDUSTRY ASSOCIATION Third Respondent PETROSA (SOC) LTD Fourth Respondent THE RETAIL MOTOR INDUSTRY ORGANISATION Fifth Respondent AMISTEC (PTY) LTD T/A LIQUID FUELS WHOLESALERS Sixth Respondent PETROLEUM RETAILERS ALIGNMENT FORUM Seventh Respondent ROYALE ENERGY Eighth Respondent NATIONAL ENERGY REGULATOR OF SOUTH AFRICA Ninth Respondent

Coram: Ingrid Opperman J

Heard: 10 November 2023

Delivered: This judgment was handed down electronically by circulation to the parties’ legal representatives by email. The date and time for hand-down is deemed to be 10h00 on 13 November 2023

Order

The application for leave to appeal is dismissed with costs including the costs of two counsel where so employed.

JUDGMENT

INGRID OPPERMAN J

Introduction

[1] This is an application for leave to appeal to the Supreme Court of Appeal or the Full Court of the Gauteng Local Division against paragraph 140.2 of this Court’s order (in terms of which the Minister’s decision was reviewed and set aside), as well as the remedial order remitting the matter to the Minister for a decision within nine months (contained in paragraph 140.3 of the order) and the costs order granted against it.

[2] This judgment should be read with the 22 September 2023 judgment (‘the main judgment’). The parties are referred to as in the main judgment and all abbreviated descriptions used herein are defined in the main judgment.

[3] The Minister and the Controller have accepted this Court’s decision. They do not seek leave to appeal against any aspect of the main judgment. They abide the Court’s decision on this application for leave to appeal.

[4] The consequence of the order granted in the main judgment is that the DoE will have to reconsider the treatment and calculation of the EC as a component of the retail margin of the RAS. That will require a review of RAS. Significantly, SAPIA has indicated that it would welcome constructive engagement with the DoE and other stakeholders in any revision of the RAS which the DoE might undertake. It has stated that it does not wish to delay any such regulatory process and contends that the appeal would not impact upon government’s right to conduct a review of the RAS, which process SAPIA would support.

[5] On its own version, SAPIA’s stated purpose in bringing this application is to “seek correction of a number of legal and factual errors in the Court’s judgment” which would “unduly fetter the Minister’s discretion in any future regulatory decisions”. The appeal is consequently not directed at this Court’s order but rather at the reasons underpinning it. That is not a recognised basis upon which an application for leave to appeal may be brought or granted. It is trite that an appeal lies against a court’s order rather than the reasons underpinning it.[1] I would dismiss this application for leave to appeal on this basis alone.

[6] However, assuming, given the facts of this case, that an appeal on the reasoning underpinning the order were competent, I would nonetheless refuse leave to appeal. In what follows I deal with the grounds of appeal and for reaching such conclusion.

Unpacking of the judgment.

[7] This court found four independent grounds upon which to impugn the Minister’s decision.

[8] Firstly, the Court found that in taking the decision to implement the RAS, the DoE and the Minister had failed properly to consider materially relevant considerations regarding the consequences of the model for retailers. The Court concluded that the decision – taken in ignorance of, or without sufficient regard to, materially relevant considerations was procedurally unfair and procedurally irrational.

[9] The Court identified a range of important considerations that were disregarded by the DoE and never placed before the Minister. The Court found that the DoE proceeded with the implementation of a pricing model based on only a selective reading of the recommendations of Bates White, that it disregarded the clear warnings raised by Bates White and that it failed to undertake the additional work proposed by Bates White. The Minister was not informed of these cautionary factors and was therefore unable to take them into account.

[10] The IPSR also raised serious concerns and flagged issues requiring additional research and analysis, but emphasised that these issues fell outside its mandate. It left the question of the EC open as it was not mandated to consider whether or not to include the EC in the retail margin. This information was before the DoE when it made the decision to implement the RAS, but there is nothing to demonstrate that the DoE applied its mind to these warnings or communicated the information to the Minister. The Minister was consequently precluded from taking these materially relevant considerations into account.

[11] The Minister was not informed of the consequences of a regulatory model that did not include a trading margin for CORO retailers. This gave rise to a failure by the Minister properly to consider the impact of the RAS on CORO retailers and to deal with the specific concerns raised by the Fuel Retailers Association.

[12] Secondly, this Court also found that the RAS model was at odds with the terms and objectives agreed by the RAS technical team. The DoE’s RAS Fact Sheet provided that the technical team had agreed that an EC should be predetermined as a reward for the retailers. In the face of this, however, the DoE adopted a methodology that is mismatched to the realities on the ground and which cannot secure the reward for retailers that the technical team agreed was required.

[13] Thirdly, this Court found that the RAS is incapable of giving effect to its objective of providing a fair return to participants in the retail sector and avoiding cross-subsidisation of activities. The decision to implement such a model is irrational to the extent that it does not ring-fence a trading margin and therefore leaves one category of participant without a fair return. The court considered, and rejected, the arguments advanced by both the DoE and SAPIA that the provision in the model for the commercial negotiation of the EC somehow remedied this structural and conceptual flaw.

[14] Lastly, this Court found that the RAS model permits and increases vertical integration within the fuel supply chain by enabling oil companies to exert an undue influence on the margins and profits of retailers and thereby to exercise control over their operations. This undermines the rationale of the RAS and closes the door on the achievement of government’s policy objectives behind the model. The implementation of a model unable to achieve its own objectives is consequently irrational.

[15] This court found that there were fundamental flaws in the decision-making process and defects in the structure of the RAS model that meant it could never achieve its objectives.

[16] SAPIA’s appeal is aimed at findings that lie at the periphery of this Court’s core conclusions. SAPIA does not deal with the fundamental findings underpinning this Court’s order.

Reliance on a RORO Site as a Benchmark Service Station

[17] SAPIA argued that this Court erred in fact and in law by impugning the use of a RORO site as the benchmark service station used in the RAS and by ultimately finding that the use of the RORO model as the benchmark service station was a basis to set aside the Minister’s decision.

[18] It is important to accurately characterise this Court’s findings on this score. This Court found that the RAS model based on a RORO benchmark station meant that CORO retailers were disregarded and would only be able to recover operational costs. This information was not communicated fully to the Minister. Although this court questioned the wisdom of basing the model solely on the RORO benchmark station, the conclusion of the Court was not that it was irrational to rely on the RORO benchmark station, but rather that the outcome of a model that did not cater for CORO stations was irrational.

[19] SAPIA’s complaint relates primarily to the findings that, under the RAS model, CORO sites cannot secure a return on investment on the operation of their retail business and that CORO site operators will inevitably be undercompensated. SAPIA argues that this Court ignored the only evidence before it being the confirmatory affidavits from retailers which – in its view – demonstrated that the retailers could sustain their businesses on the current model. SAPIA’s complaints overlook the fact that this evidence was considered, but that the defect with the RAS model was a conceptual one which went beyond the factual

question of the viability of specific retailers.

[20] The judgment provides that the model cannot be designed to, in principle, not provide an EC to the retailer at all. The RAS model compensates retailers for operational costs, but does not secure a trading margin for them. Profits are realised through the CAPEX margin, which the RAS contemplates will be earned by the owners of the capital assets at a service station. Where

the retailer does not own the assets, the RAS does not provide for the retailer to earn a profit at all. The RAS requires the owner of the assets (the wholesaler in the case of a RORO station) to forfeit a portion of the profits to which it would otherwise be entitled (from the CAPEX margin). It was that conceptual flaw in the model that led this court to ultimately find that the RAS could not achieve its objectives. This Court found that to deny the structural flaw by pointing to instances where the structural flaw has not manifested itself particularly harshly is to deny the inherent and logical flaw in the model on anecdotal evidence. What was required, was to consider the scheme in its full context.

[21] No factual findings on the overall profitability or viability of retailers were made. The Court’s conclusion was a far narrower one, namely that retailers were undercompensated for the risk they assume when running a business of a petrol station, and for the fuel-dispensing service they provide. The scheme does not provide a mechanism for retailers who do not own capital assets to realise any profits. Their profits depend on the outcome of negotiations with retailers in circumstances where wholesalers are expected to “forfeit” a portion of the profits the RAS model assumes they should receive as a result of their ownership of the relevant assets.

[22] This scheme is flawed; a conclusion independent of the factual question of the profitability of the retailers. SAPIA’s complaints with the judgment do not address these fundamental flaws in the RAS model.

Unequal bargaining power

[23] SAPIA complains that this Court erred in finding that every fuel retailer was at an automatic commercial disadvantage in negotiations with the oil companies. This was not the finding. The Court accepted that there was unequal bargaining power between the fuel retailer and the oil company in the context of any negotiation about the EC. This was based on the statements of the Constitutional Court in Rissik Street One Stop CC t/a Rissik Street Engen and Another v Engen Petroleum[2] as well as the provisions and preamble of the Petroleum Products Act. There was no blanket finding in relation to commercial disadvantage,

but rather an acknowledgement of the unequal bargaining power inherent in the petroleum industry. The Court’s finding

also recognised that the ‘default position’ in the RAS model that allocated the EC to the asset-owner would further skew the bargaining position of both parties as the retailer either accepts what is on offer from the oil company, or it must walk away from the transaction.

[24] SAPIA’s attack on the findings regarding the bargaining power between the oil companies and the fuel retailers does not displace the key finding in the judgment on this score which was that even if there were no imbalance in bargaining power, the inevitable outcome of the RAS model was that either the retailers were undercompensated, or the asset-owners were undercompensated. This is the fundamental flaw in the RAS, which rendered it arbitrary and incapable of giving effect to the objectives that it was designed to achieve. SAPIA’s appeal does not address or attack the findings regarding this flaw in the RAS.

[25] SAPIA’s attack does not displace the key findings in the judgment on this issue as even if there were no imbalance in bargaining power, the inevitable outcome of the RAS model was that either the retailers were undercompensated, or the asset-owners were undercompensated. SAPIA’s appeal does not address or attack the Court’s core findings regarding this flaw in the RAS.

Regulation of the Entrepeneurial Compensation

[26] SAPIA contends that this Court found that the EC should be regulated and not left to negotiation between the fuel retailers and the oil companies. This is not so. This court found that the treatment and allocation of the EC for retailers in CORO sites as an allocation within the retail margin of the RAS is referred back to the DoE for reconsideration. The judgment envisages further consideration of the model in its entirety as it relates to CORO stations to ensure that any model can produce outcomes in light with its objectives. It remains for the DoE and the Minister to determine how to achieve the objective of ensuring fair

and transparent returns for each activity in the fuel supply chain, as well as the other purposes of the RAS (including certainty and the prevention of vertical integration).

[27] The Court explained that the RAS does not provide for an EC. The Guidelines, Matrices or Principles don’t entitle the retailers to claim an EC either. Rather the model leaves it to the investor and retailer to decide what portion of the CAPEX margin the investor will forfeit to the retailer. It is SAPIA’s members (the oil companies) who are entitled – on paper –

to the full CAPEX portion of the retail margin, and who are expected to “forfeit” a portion of this margin to the retailers.

The retailers are forced to operate within a regulatory scheme that makes no provision at all for a secured profit margin on their

investment in the business of operating a service station. It is not the fact that the model leaves certain portions of the margin to negotiation that is irrational per se. It is the fact that this negotiation is expected to occur in a context where the negotiating power is unequal, the default position is that the oil company must “forfeit” a portion of the capex margin, and no negotiating guidelines have been put in place.

[28] The need for a regulated EC is not based entirely on the fate of the retailers. The model gives rise to perverse outcomes for all parties. The model is structured so that either the retailers are undercompensated for the risk of operating the petrol business, or the asset owners are undercompensated for their investment in the assets.

[29] SAPIA’s appeal does not address or challenge these fundamental underpinnings of this Court’s judgment.

The Minister’s decision-making process

[30] SAPIA contends that this Court found the decision-making process unlawful because the Minister did not adopt unaltered the input received from industry experts and that the Bates White report should dictate the Minister’s decision (thereby suggesting that the Court found that there was an obligation on the Minister to follow the Bates White recommendations). SAPIA has mischaracterised the findings on both scores. The judgment does not state that the Minister ought to have followed the input of industry experts and the Bates White Report (and her decision was flawed on the basis that she failed to do so). Rather, the judgment provides that the reviewable errors in the decision-making process arose from the fact that the Minister was not properly appraised of the contents of the Bates White report and therefore did not take it into account. This was also the case in respect of the opinion and recommendations in the IPSR report. SAPIA has not pointed to any evidence in the papers or record of decision that demonstrates that the Minister did in fact apply her mind to this information.

[31] SAPIA contends that the Court found that the use of a summary by the DoE in the recommendations to the Minister was in and of itself a reviewable error. That is not what the judgment says. The issue was that this summary excluded material information, provisos and warnings, as well as recommendations that further work be completed before a model was adopted. This prevented the Minister from taking these considerations into account at all.

[32] SAPIA contends that this court found in effect that the Minister was not entitled to disagree with certain inputs from Bates White, IPSR and the Fuel Retailers Association. This is incorrect. The reviewable irregularity in the process was a failure to consider these inputs at all, and not the fact that the DoE or the Minister took a different view on these important issues.

[33] The Rule 53 Record demonstrated that the Minister received inaccurately summarised information that did not include a proper explanation of the consequences of the implementation of the recommended model.

Vertical Integration

[34] SAPIA’s final ground of appeal is that the Court was wrong to conclude that the negotiation of the EC within the current RAS model permits vertical integration. SAPIA provides no further explanation for why the reasoning and findings of the Court should be set aside on appeal. I can see none.

Conclusion

[35] In the decision of Dexgroup (Pty) Ltd v Trustco Group International (Pty) Ltd and Others[3], Wallis JA observed that a court should not grant leave to appeal, and indeed is under a duty not to do so, where the threshold which warrants such leave, has not been cleared by an applicant in an application for leave to appeal. In paragraph [24] he held as follows:

“[24] For those reasons the court below was correct to dismiss the challenge to the arbitrator's award and the appeal must fail. I should however mention that the learned acting judge did not give any reasons for granting leave to appeal. This is unfortunate as it left us in the dark as to her reasons for thinking that enjoyed reasonable prospects of success. Clearly it did not. Although points of some interest in arbitration law have been canvassed in this judgment, they would have arisen on some other occasion and, as has been demonstrated, the appeal was bound to fail on the facts. The need to obtain leave to appeal is a valuable tool in ensuring that scarce judicial resources are not spent on appeals that lack merit. It should in this case have been deployed by refusing leave to appeal.” (emphasis added)

[36] I have considered the extensive application for leave to appeal dispassionately[4] and hold the view that the appeal would not have a reasonable prospect of success nor that there is some other compelling reason

why the appeal should be heard. Most of the grounds have been answered in the main judgment. Nothing argued has persuaded me that

another court would find differently.

[37] The criticism that the Minister’s discretion will be hamstrung by the judgment ie that her discretion will be fettered, in my view is without merit having regard to the actual findings made in the judgment.

[38] As mentioned, SAPIA has indicated that it would welcome constructive engagement with the DoE and other stakeholders in any revision of the RAS which the DoE might undertake. It has stated that it does not wish to delay any such regulatory process and contends that the appeal would not impact upon government’s right to conduct a review of the RAS, which SAPIA would support. The position SAPIA holds appears to be contradictory: either the reasoning in the judgment is flawed in which event it must be corrected before a revision is embarked upon or the stakeholders can engage on the revision process forthwith which implies that a process can be undertaken having regard to the findings in the judgment which will not fetter the Minister’s discretion in any future regulatory decisions unduly. If it is accepted that the Minister’s discretion is unfettered then there is nothing left to correct on appeal and the application under consideration falls to be dismissed.

[39] I accordingly grant the following order:

The application for leave to appeal is dismissed with costs to include the costs consequent upon the employment of two counsel, where so employed.

I OPPERMAN

Judge of the High Court

Gauteng Division, Johannesburg

Counsel for the Applicant (SAPIA):

Adv MM le Roux SC

Instructed by:

Fasken Attorneys

Counsel for the Respondent (Fuel Retailers Assocation):

Adv G Quixley and Adv F Hobden

Seton Smith & Associates

Date of hearing: 10 November 2023

Date of Judgment: 13 November 2023

[1] Western Johannesburg Rent Board & another v Ursula Mansions (Pty) Ltd 1948 (3) SA 353 (A) at 355; ABSA Bank Ltd v Mkhize and two similar cases 2014 (5) SA 16 (SCA) para 64

[2] Rissik Street One Stop CC t/a Rissik Street Engen and Another V Engen Petroleum 2023 (4) BCLR 425 (CC).

[3] 2013 (6) SA 520 (SCA)

[4] Smith v S, 2012 (1) SACR 567 (SCA)