2
installation of infrastructure, known as “daughter stations” at each of the
hospitals so as to enable the gas piped to the hospitals to be stored and then
pumped into the furnaces.
2 On 22 September 2015, GDID contracted with VGN for installation of daughter
stations at, and the supply of natural gas to, four Gauteng hospitals for a period
of three years. The gas was supplied and priced on the basis that GDID would
pay for a minimum amount of gas to be made available to each of the
hospitals, even if the hospitals did not use it. This was referred to in the
contract as the “take or pay” principle. The clearest delineation of that principle
emerges in paragraph 10.4 of the September 2015 agreement, which states
that GDID “shall be obliged to, in each contract year, purchase, take or pay
for” an “adjusted annual contract quantity”. That quantity is reached by taking
the “annual contract quantity” and varying it by a method set out in the
agreement which is not important for present purposes. The annual contract
quantity was just under 410 000 gigajoules of gas.
3 However, it turned out that the four hospitals covered by the September 2015
contract did not use anything close to the volume of gas VGN tendered to
supply, and GDID promised to pay for.
4 On 29 November 2016, therefore, the parties entered into an addendum,
which extended the term of the initial contract, and expanded its scope. The
idea appears to have been to enhance the capacity of Gauteng’s health
infrastructure to the point at which it would be able to use the gas it had agreed
to pay for. The addendum provided for the supply of natural gas to an
additional six hospitals, together with the provision of infrastructure (in the form
3
of gas furnaces and daughter stations) to those hospitals to allow them to use
the gas supplied. The period of the September 2015 agreement was extended
by two years in relation to the four hospitals to which it applied. In relation to
the other six hospitals, the duration of the contract for the supply of natural
gas was specified as five years from the date on which certificates of
completion of the enhanced infrastructure necessary to be installed at those
hospitals were signed (see clause 4.1 of the addendum). VGN obtained a loan
of R39 million from the Industrial Development Corporation to fund the
construction of the additional infrastructure. Where I refer to the “the
agreement” or “the contract” elsewhere in this judgment, I mean the
September 2015 agreement and the addendum, read together.
5 Throughout this period, GDID regularly defaulted on its payment obligations.
VGN obtained arbitral awards enforcing these obligations. The amounts due
under the awards have now been paid, but the second award is of significance
because it defined what VGN refers to as GDID’s “primary obligation” under
both the original and the amended contract: that obligation, VGN says, is to
pay, annually, for the use of just under 410 000 gigajoules of gas throughout
the duration of the contract – whether or not that gas is actually used at any
of the hospitals.
6 On 19 January 2023, GDID gave notice that it regarded the agreement as
having expired in relation to six of the ten hospitals, and having either “not
expired” or “not commenced” in relation to the other four. Though what exactly
was meant by “expired”, “not expired” and “not commenced” is not clear on
4
the face of the 19 January 2023 notice, the final paragraph of the notice was
unambiguous, in that it purported to terminate the whole agreement.
7 VGN did not accept that GDID was entitled to terminate the agreement. It
regarded the agreement as not having run its course until five years after the
additional infrastructure to be installed at the six hospitals covered by the
addendum had been completed. VGN saw the 19 January 2023 termination
as an unlawful repudiation. It accepted that repudiation and now claims
contractual damages in the form of unpaid rental for daughter stations installed
at certain of the hospitals, lost past and future profits resulting from GDID’s
failure to honour the contract, and interest on those amounts. VGN also seeks
to recoup the value of interest payments it had to make on its IDC loan that
would not have been due had GDID timeously satisfied an arbitral award VGN
obtained for the payment of amounts due for gas supplied under the
agreement.
Was the contract repudiated or terminated?
8 Much of the argument before me concerned the dispute about the impact the
addendum had on the duration of the agreement. Clause 4.1 of the addendum
states as follows –
“The period for the initial agreement is extended by a further 2 (two) years[‘]
period and the additional sites will run for a period of 5 (five) years
commencing from signing of the completion certificates”.
9 VGN says that the clause means what it says: the period of the agreement in
respect of the four hospitals named in the September 2015 agreement was
extended to five years, and the period of the agreement applicable to the six
5
hospitals named in the addendum was to run until five years after GDID
certified that the infrastructure had been installed at the last of the six hospitals
named in the addendum. Throughout this time, VGN contends, GDID
assumed the “primary obligation” to purchase, annually, just under 410 000
gigajoules of gas (adjusted where necessary in line with the provisions of the
September 2015 agreement).
10 GDID contends that the agreement subsisted for a maximum of five years from
the date on which the addendum was entered into – that is, no later than 30
November 2021. The termination of the agreement on 19 January 2023 was
accordingly no more than the affirmation of a state of affairs that came into
being when the agreement expired by effluxion of time.
11 In my view, GDID’s interpretation of the effect of the addendum is untenable.
It is in the first place inconsistent with the plain text of clause 4.1. But it also
makes no sense in the context of the agreement as a whole, evaluated in its
surrounding circumstances. In order to accept GDID’s interpretation, I would
have to accept that the duration of the extended agreement was a fixed five-
year period, and that it started to run on the day the addendum was signed.
12 But that makes no business sense. The sensible interpretation is that the
duration of the extended agreement was triggered on the day the addendum
was signed, but that this duration was an indefinite but ascertainable period .
That period is equal to the length of time between the signature of the
addendum and the certification of the completed infrastructure work on the
last of the hospitals named in the addendum, plus five years. The underlying
scheme of the agreement was clearly that VGN would place each of the ten
6
named hospitals in the position to use gas-fired furnaces for a period of five
years, and that it would supply gas to GDID for use at the hospitals until five
years after the last of the hospitals was placed in such a position.
13 GDID advanced no coherent argument to the contrary. I reject its contention
that the agreement expired five years from the date on which the addendum
was signed. It in fact seems to me that GDID’s reference, in its notice of
termination, to obligations under the agreement in respect of some of the
hospitals having “not commenced” is a tacit acceptance that the agreement
had a much longer shelf life than a fixed five year period. There is no indication
in the text of the agreement, in its avowed or imputed purpose, or in the
circumstances surrounding its conclusion, that GDID’s rights and obligations
in relation to each of the six hospitals named in the addendum were severable
from one another, with the effect that GDID could resile from the agreement
after work on one hospital had been completed but before work on another
had started. The better view is that GDID’s obligations under the agreement
were indivisible – at least in relation to the six hospitals named in the
addendum. As a consequence, the duration of the agreement continued until
five years after the last of the work had been done and certified on the last of
the hospitals.
14 It follows that GDID was not entitled to terminate the contract when it did, and
that its notice of termination was, in fact, a repudiation of the agreement.
Was the decision to enter into the addendum unlawful?
15 In a counter-application, GDID contends that, if the duration of the agreement
is as VGN alleges, then GDID’s decision to enter into the agreement was
7
unlawful. It asks for an order reviewing and setting aside that decision. GDID
also seeks certain other relief in the counter-application, including an order
varying one of the arbitration awards, and an order declaring VGN to have
been overpaid by GDID in the sum of almost R79 million. Mr. Khoza, who
appeared with Mr. Chavalala for GDID, confirmed that GDID no longer persists
with the application to vary the arbitration awards. Although GDID continues
to press for the overpayment relief, I need not rule on that part of the counter-
application, in light of the view I have taken on the determination of the
quantum of VGN’s claim.
16 The review was advanced on two bases. The first was that the addendum was
unlawfully entered into because it expanded the scope of work awarded to
VGN beyond the terms of the original request for proposals. The second was
that the addendum had the effect of increasing the value of VGN’s contract
beyond the margin approved in a National Treasury Instruction said to be
binding upon GDID. I address each point in turn.
The scope of the request for proposals
17 GDID has not been consistent in its submissions on this score. At paragraphs
8.1. 8.1.1 and 8.1.18 of its answering affidavit, GDID alleges that the initial
request for proposals (RPF07/07/2014) invited bids in respect of the work
covered in the addendum, and that this work was in fact awarded to VGN.
However, the documents annexed to the answering affidavit do not support
the inference that the scope of the work for which VGN originally tendered and
for which it was appointed included the work identified in the addendum. The
request for proposals itself is not annexed to the answering affidavit. The
8
documents that are annexed appear to confirm that the scope of work for
which VGN initially tendered, and for which it was appointed, was limited to
four hospitals, none of which is referred to in the addendum.
18 I queried this contradiction at the hearing of this matter convened on 15 May
2026. Mr. Louw, who appeared for VGN, contended that the situation was in
fact as was described in paragraph 8 of GDID’s answering affidavit: viz. that
the request for proposals covered all the work for which VGN was ultimately
contracted. This drew no response from Mr. Khoza.
19 Another hearing was held on 17 August 2026 to canvass further argument on
GDID’s review application, and to argue an application for leave to amend
GDID’s notice of motion. During that hearing, Mr. Khoza asserted that the
request for proposals did not encompass the work covered in the addendum,
and that the addendum unlawfully widened the scope of the work for that
reason. However, despite being given a post-hearing opportunity to say where
the request for proposals appears on the papers, or to apply for leave to
introduce it, counsel for GDID did neither. Their case was in fact that the
request for proposals appeared at annexure “D2” to the answering affidavit.
But that was plainly erroneous. Annexure “D2” is in fact VGN’s response to
GDID’s invitation to bid. It does not define the scope of the work for which VGN
was invited to bid.
20 For his part, Mr. Louw argued in his post-hearing submissions that GDID must
stand or fall by the version set out at paragraphs 8.1. 8.1.1 and 8.1.18 of its
answering affidavit: viz. that the request for proposals did include the six
hospitals identified in the addendum. Given the stance GDID has taken in this
9
case, I must agree. Notwithstanding the apparent tension between the version
put in GDID’s answering affidavit, and the annexures upon which it relies,
GDID must be held to the version set out in the answering affidavit. On that
version, the hospitals named in the addendum were within the scope of the
work defined in the request for proposals.
21 The argument that the addendum exceeded the scope of that work must
accordingly be rejected.
The treasury instruction
22 The treasury instruction was introduced to supplement GDID’s review after the
hearing of 15 May 2026. Understandably, VGN objected to its late introduction
and to GDID’s reliance upon it. On balance, given that GDID’s case is
fundamentally one of legality, it is in the interests of justice that I consider the
instruction, since to do otherwise would risk adjudicating the case on the basis
of a legal misconception.
23 In any event, I do not think the treasury instruction assists GDID’s case. This
is so for two reasons. The first is that, even assuming that the treasury
instruction forbade GDID from entering into the addendum, the instruction
itself cannot have the status of law. It is plainly an internal administrative
document produced by a treasury official with no law-making powers. It was
not published in the government gazette, and accordingly lacks the publicity
necessary to ground a review application. In my view, legality review can only
be sustained on the basis of publicly accessible legal rules which have been
promulgated in accordance with a specific empowering provision. The
treasury instruction lacks either of these attributes, and is, accordingly, not
10
“law” in the sense that can sustain a legality review of the kind GDID seeks to
pursue.
24 To hold otherwise would risk blurring the distinction between law on the one
hand, and policy or internal administrative practice on the other. Mr. Khoza
argued that GDID is bound by the treasury instruction because section 76 of
the Public Finance Management Act 1 of 1999 says so. That is true, but it does
not follow that GDID is entitled to self-review its decision to enter into a
contract in breach of one of those instructions. Self-review is a form of legality
review. Legality review requires the reviewing party to show that the act being
reviewed breaches a law. For the reasons I have already given, the treasury
instruction is not “law” in this sense. It certainly binds GDID in its dealings with
the treasury, but it does not entitle GDID to impugn its own conduct toward
third parties, who cannot be expected to be aware of the instruction in
advance, precisely because the instruction lacks the publicity that a
promulgated law would have.
25 In any event, I do not think that it has been established that the decision to
enter into the addendum was taken in breach of the treasury instruction. The
treasury instruction forbids more than a 15% variation in the value of a
government contract. For the reasons I turn to below, it has not been
established that the addendum had this effect.
26 Accordingly, the counter-review must be dismissed. I have reached this
conclusion without considering whether GDID’s self-review was unreasonably
delayed, and whether it would have been appropriate to overlook any such
delay in this case. Given the transparent lack of merit in the review, I would
11
have been hard-pressed to overlook the substantial delay in this case, but,
because everything turned on GDID’s prospects of success, it was in the
interests of justice that I consider the review on its merits.
The quantum of VGN’s damage
27 It follows from all this that GDID repudiated the agreement, and that VGN is
entitled to be compensated for such damage as it may prove in consequence
of that repudiation. The quantum of damage VGN claims on the papers before
me depends on my accepting that the “primary obligation” to take or pay for
just under 410 000 gigajoules of gas annually survived the parties’ agreement
to the addendum. Whether that is so is conceptually and practically distinct
from the question of the duration of the agreement as a whole. If the
agreement was supposed to subsist until five years from the installation of new
infrastructure at the last of the hospitals named in the addendum, then it is
clear that VGN is owed something in respect of the period between GDID’s
repudiation of the agreement and the point at which the contract would
otherwise have come to an end. However, it does not mean that VGN is
entitled to the quantum it claims in this application, since that quantum is
calculated on the assumption that the “primary obligation” to take or pay for
just under 410 000 gigajoules of gas annually survived the addendum , and
that this “primary obligation” was meant to apply throughout the duration of the
agreement.
28 The problem for VGN is that clause 5.1 of the addendum says this –
“5.1. The Parties agree that the take or pay principle as agreed upon in the
[September 2015] Agreement as well as any reference to "Minimum
Order Quantities" shall no longer be applicable and that [GDID] shall be
12
charged monthly the product and or compressed natural gas [sic]
delivered for that specific month and shall only pay for the compressed
natural gas delivered for that month.”
29 On its face, clause 5.1 of the addendum suggests that neither the obligation
to purchase the adjusted annual contract quantity nor the “take or pay
principle” entrenched in the initial agreement applies any longer to the overall
agreement between the parties. In its answering affidavit, GDID is emphatic
that this is exactly what the clause means (see especially paragraphs 180,
185, 187 and 192 of the answering affidavit).
30 VGN contends otherwise. It says that the effect of the clause was that “it did
away with the monthly take or pay principle. But it did not amend the total
annual [compressed natural gas] GDID had to order” (see paragraph 47 of
VGN’s founding affidavit. The emphasis is in the original). VGN also relies on
several other provisions of the September 2015 agreement which indicate that
GDID’s primary obligation under both the main agreement and the addendum
was to “purchase, take or pay” a minimum amount annually (see paragraphs
74 to 80 of the founding affidavit). VGN furthermore points out that, unless
GDID was obliged to “purchase, take or pay for” a minimum amount of gas
annually, the agreement would lack business efficacy, since VGN was bound
to pay its own suppliers on the same “take or pay” principle. VGN also relies
upon an arbitral award it secured in which it was confirmed, so VGN contends,
that GDID’s primary obligation was to “purchase, take or pay” just under
410 000 gigajoules of gas annually. VGN says that award renders the issue
res judicata. These arguments make sense on their own terms, but they do
not accord with the plain text of clause 5.1 of the addendum.
13
31 Perhaps because GDID changed its counsel shortly before the hearing , the
fundamental difficulty that arises from clause 5.1 of the addendum was
overlooked. Understandably, VGN’s papers gloss over that difficulty by
pointing to terms of the September 2015 agreement and the surrounding
circumstances which suggest that the take or pay principle was fundamental
to the contract, and to the arbitrator’s award which appears to adopt the view
that the take or pay principle survived the addendum.
32 Less intelligible is GDID’s counsels’ lack of emphasis on difficulties clause 5.1
causes for the calculation of VGN’s damages. This is despite clear and
emphatic reliance on the text clause 5.1 in GDID’s answering affidavit , to
which VGN offers no specific reply.
33 In my view, however, the meaning of clause 5.1 is central to determining the
value of VGN’s claim. Fundamental to VGN’s case is that GDID’s primary
obligation was to purchase a minimum amount of compressed natural gas for
the duration of the contract between the parties. Indeed, it seems to me that
the agreement may lack much of its business efficacy unless that obligation
survived the addendum. On the other hand, clause 5.1 of the addendum
clearly states that the take or pay principle was to be abandoned, and that
GDID would henceforth only pay monthly for the amounts of gas actually
“delivered” in a specific month.
34 I do not accept that I am bound by the second arbitral award on this question.
That award was made on an unopposed basis, and did no more than accept
VGN’s submissions as to the proper interpretation of clause 5.1 of the
addendum. Those submissions have never squarely addressed the textual
14
problems raised by clause 5.1. They either skirt around or seek to avoid them
by reference to other provisions of the agreement, or to the second arbitral
award. In these circumstances, GDID is right to submit, in its answering
affidavit, that it is entitled to challenge VGN’s submissions in these
proceedings. To hold otherwise would require me to conclude that GDID is
issue estopped from re-opening any controversy about meaning of clause 5.1
of the addendum. Given the amounts of public money involved, VGN’s failure
to face up to the problems that the text of clause 5.1 causes, and the fact that
GDID has never been heard on this issue, I think that result would be
inequitable (see in this respect Prinsloo NO v Goldex 15 (Pty) Ltd 2014 (5) SA
297 (SCA), paragraph 26).
35 It is by now trite that the interpretation of contracts is a unitary exercise which
takes in to account the words used in a written agreement, the underlying
purpose of that agreement, and the circumstances surrounding the striking of
the agreement . Courts will generally be keen to adopt interpretations of
contracts which have sensible and businesslike results (see Natal Joint
Municipal Pension Fund v Endumeni Municipality 2012 (4) SA 593 (SCA) at
paragraph 18).
36 It seems to me that the plain text of clause 5.1 of the agreement is apparently
incongruous with the underlying purpose and business efficacy of the
agreement. This is not a case where the purpose and effect of the clause can
be inferred from the type of contract at issue or from some special feature of
the text and its structure (Capitec Bank Holdings Limited v Coral Lagoon
Investments 194 (Pty) Ltd 2022 (1) SA 100 (SCA) paragraphs 50 and 51). The
15
agreement at issue in this case is a bespoke arrangement, meant to achieve
a unique purpose. The proper interpretation of clause 5.1 against the
background of the agreement as a whole requires an unusually sensitive
contextual treatment. At present, the indications of meaning provided by text
of the agreement on the one hand, and by the purpose and surrounding
circumstances of the agreement on the other, are evenly balanced. Its true
meaning can only be determined by a fuller investigation of the circumstances
surrounding the conclusion of the addendum – in other words, by oral
evidence of those circumstances.
37 Mr. Louw submitted that if, for any reason, VGN’s damages could not be
determined on affidavit, the appropriate relief would be to declare that the
agreement was repudiated and to refer the quantification of VGN’s damages
to trial. I accept that submission.
Remedy
38 It follows from all of this that –
38.1 The addendum entered into in November 2016 extended the
duration of the contract to an indefinite b ut ascertainable date not
later than five years after GDID certified that VGN had given the last
of the hospitals named in the addendum the infrastructure necessary
to use gas-fired furnaces with gas VGN supplied. GDID’s termination
of the agreement before that date was in fact a repudiation of GDID’s
obligations under it.
16
38.2 The decision to enter into the addendum was not unlawful on either
of the bases GDID has identified. The addendum did not unlawfully
exceed the scope of the work GDID invited VGN to tender for. The
treasury instruction upon which GDID relies does not have the force
of law, and accordingly cannot be relied upon in GDID’s self-review.
It is, in any event, impossible to determine whether the addendum
varied the value of the contract in breach of the treasury instruction
upon which GDID relies, because it is presently impossible to
determine, without the benefit of oral evidence, whether GDID’s
obligation to purchase a minimum amount of gas from VGN survived
the addendum. The review application must fail.
38.3 VGN is in principle entitled to such damages as it may prove flow
from GDID’s repudiation of the agreement. However, the quantum of
those damages depends, in part, on whether GDID’s obligation to
purchase a minimum amount of gas from VGN survived the
addendum. That question must be referred to trial, together with the
balance of the issues relevant to the determination of the quantum of
VGN’s damages.
Costs
39 VGN has been successful on every issue finally determined in this judgment.
Costs must follow that result.
Order
40 Accordingly –
17
40.1 The application for leave to amend the respondent’s notice of
counter-application is granted.
40.2 The parties’ respective applications for leave to file further written
submissions are granted.
40.3 Prayers 1, 2, 4 and 5 of the counter-application are dismissed.
40.4 Prayer 3 of the counter-application is postponed sine die, to be
pursued together with the quantification of the applicant’s contractual
damages in terms of paragraphs 40.6 and 40.7 below.
40.5 It is declared that -
40.5.1 the respondent breached its contract with the applicant
(comprising both the main agreement of 22 September
2015 and the addendum of 29 November 2016) by
purporting to terminate the contract when the respondent
had no right to do so, thereby repudiating the contract; and
that
40.5.2 the applicant is entitled to such damages as it may prove
flowing from this breach.
40.6 The question of the quantum of the applicant’s damages is referred
to trial. The applicant’s notice of motion and founding affidavit will
stand as its combined summons. The answering affidavit will stand
as the respondent’s plea. The replying affidavit will stand as the