CONFIDENTIAL

REPUBLIC OF MOZAMBIQUE PIPELINE INVESTMENTS

COMPANY (PTY) LTD

NERSA TARIFF APPLICATION FOR THE ADDITIONAL

VOLUMES TRANSPORTED UNDER THE SECOND GAS

TRANSPORT AGREEMENT (GTA2)

31 July 2012

Table of Contents

1 Introduction ..................................................................................................................... 1

2 Background ..................................................................................................................... 1

2.1 Transmission Agreements ....................................................................................... 2

2.2 The GTA1 and GTA2 tariffs ..................................................................................... 3

3 Methodology ................................................................................................................... 3

3.1 Methodology selection: Discounted cash flow (Levelised cost) ................................ 3

3.2 Levelised cost implementation: the details ............................................................... 5

4 Selection of the system cost basis .................................................................................. 6

4.1 Framework for selecting the system cost basis ........................................................ 7

4.2 The implications for the GTA2 tariff .......................................................................... 8

4.3 An alternative view ................................................................................................. 10

4.4 System cost basis: conclusions.............................................................................. 10

5 The cash flows .............................................................................................................. 11

5.1 Capital expenditure ................................................................................................ 12

5.2 Operating costs ...................................................................................................... 12

5.3 Debt finance........................................................................................................... 13

5.4 Tax ........................................................................................................................ 14

6 Discount rate ................................................................................................................. 14

6.1 Beta ....................................................................................................................... 15

6.2 Risk Free Rate and Market Risk Premium ............................................................. 17

6.3 Small Stock Premium ............................................................................................. 17

6.4 Second Country Risk Premium .............................................................................. 18

7 Volume forecasts .......................................................................................................... 18

8 Tariff calculation ............................................................................................................ 19

9 Conclusions .................................................................................................................. 20

Appendix A: Determination of the Equity Beta for ROMPCO ................................................ 21

Appendix B: Detailed tables ................................................................................................. 22

Appendix C: Updated NERSA risk free rate and market risk premium spreadsheet ............. 29

3

Tables Table 1: Capital expenditure ............................................................................................... 12

Table 2: Total operating costs ............................................................................................. 12

Table 3: Cost of debt ........................................................................................................... 14

Table 4: Second country risk adjustment ............................................................................. 18

Table 5: Levelised cost calculation ...................................................................................... 19

Table 6: Fixed operating costs ............................................................................................ 22

Table 7: Components of the cost of equity calculation......................................................... 22

Table 8: Details of cost of equity calculation (1) .................................................................. 23

Table 9: Details of cost of equity calculation (2) .................................................................. 23

Table 10: Debt related cash flows ....................................................................................... 24

Table 11: Historic and forecast volumes.............................................................................. 25

Table 12: Tax treatment of assets ....................................................................................... 26

Table 13: Tax calculation .................................................................................................... 27

Table 14: Discounted cash flow calculation ......................................................................... 28

Figures Figure 1: Diagrammatic representation of the pipeline network ............................................. 2

Figure 2: Small Stock Premium applied in South Africa ....................................................... 17

1

1 Introduction

1. This document contains the Republic of Mozambique Pipeline Investments Company

(Proprietary) Limited„s (ROMPCO) application to the National Energy Regulator of South

Africa (NERSA) for approval of the tariff to be charged under its General Gas Transport

Agreement Number Two (GTA2). The tariff has been calculated in terms of the guidelines

document published by NERSA, entitled: “Guidelines for Monitoring and Approving Piped-

Gas Transmission and Storage Tariffs in South Africa.”1, (“the NERSA Guidelines”).

2 Background

2. ROMPCO is the owner of the 26 inch 869 km high pressure cross border gas transmission

pipeline from Mozambique to Secunda in South Africa, also known as the Mozambique to

Secunda Pipeline (MSP), shown in Figure 1 below. The pipeline transports gas produced

from gas fields in the vicinity of Vilancoulus in Mozambique to Secunda in South Africa

where the gas is delivered into the South African gas transmission and distribution pipeline

network owned and operated by Sasol Gas Limited. The pipeline facility owned by ROMPCO

includes a Pressure Protection Station (PPS) situated in Secunda. A small amount of gas is

also delivered at Ressano Garcia in Mozambique.

3. ROMPCO is a joint venture company with three shareholders, namely the South African Gas

Development Company (Proprietary) Limited (iGas) (25%), Companhia Moçambicana de

Gasoduto S.A.R.L (CMG) (25%) and Sasol Gas Holdings (Proprietary) Limited (50%).

Pursuant to agreements with the South African Government and the Mozambican

Government respectively, iGas and CMG were nominated by the respective governments as

the designated shareholders in the company.

4. In accordance with the agreements with the two governments, ROMPCO appointed Sasol

Gas Limited as the Operator of the Pipeline to operate and maintain the MSP pipeline that

was commissioned at the end of March 2004, and entered into an agreement in this regard

otherwise known as the Operating and Maintenance Agreement (OMA).

1 Dated 1 May 2009.

2

Figure 1: Diagrammatic representation of the pipeline network

2.1 Transmission Agreements

5. A Gas Transportation Agreement (GTA1) was signed with Sasol Gas in December 2002 for

the transportation of 120 MGJ/a of gas to Secunda. In July 2006, ROMPCO also entered

into an agreement with the Government of Mozambique, represented by the National

Petroleum Institute (“INP”), to transport 6 MGJ/a of royalty gas at a preferred tariff in

Mozambique. Currently, 3 MGJ/a of this is delivered in Mozambique.

6. Subsequently, the need arose to transport a further 27 MGJ/a to Secunda. ROMPCO had to

expand the MSP capacity in order to meet this need. On 17 May 2007 the ROMPCO Board

approved expenditure of R1,06 billion to install a Compressor Station at Komatipoort. This

was subject to the signing of a Heads of Agreement with Sasol Gas for delivery of an

incremental 27 MGJ/a to Secunda. This Term Sheet (GTA2) was signed in September 2007.

The tariff therein underpinned the return on investment on the Compressor investment by

ROMPCO.

AE

B

C

D

Petronet PipelineROMPCO PipelineSasol Gas Pipelines

Produced by:

SASOL GAS GIS TEAMAlrodeMarch 2006

REV 001

Pretoria

Benoni

Nigel

ErmeloBethalAlberton

Rosslyn

Secunda

Witbank

BreytenLenasia

Springs

Babelegi

Hendrina

Badplaas

Malelane

Carolina

Meyerton

Barberton

Nelspruit

Volksrust

Amsterdam

Sasolburg

Lyttelton

Newcastle

Middelburg

Standerton

Kemptonpark

Komatipoort

Vereeniging

Randfontein

Phoenix

Richardsbay

Avoca

Durban

Verulam

Mandini

Empangeni

Ressano Garcia

Maputo

Sabie

Guija

Temane

Motaze

Mahele Chokwe

Muabsa

Talofo

Jofane

Chigubo

Panjane

Meginge

Componde

Magandene

Camo-Camo

Rumbacaca

Vilanculos

Mabuiapanse

MPUMALANGA

GAUTENG

KWAZULU-NATAL

MAPUTO

MAGUDE

GAZA

INHAMBANE

MOAMBA

SWAZILAND

MOZAMBIQUE

SOUTH AFRICA

Germiston: Operations

126 (*80) employees

Randburg: Head Office

65 (*65) employees

Nelspruit: Depot

8 (*8) employees

Mozambique: Depot

7 (*7) employees

Secunda: Control room

5 (*5) employees

Pinetown: Depot

8 (*8) employees

3

2.2 The GTA1 and GTA2 tariffs

7. The GTA1 tariff is not subject to NERSA approval, as confirmed during discussions with

NERSA early in 2012. This is established in the agreement relating to the MSP between

Sasol and the Government of South Africa which is incorporated into Section 36 of the Gas

Act, 2001 (Act No. 48 of 2001, “the Act”). This tariff is furthermore confirmed in the license

conditions of the transmission license issued by NERSA to ROMPCO in terms of Section 15

of the Gas Act.

8. On 23 August 2011 ROMPCO applied to NERSA for approval of the transmission tariff

currently in place for GTA2 in terms of the Term Sheet signed in September 2007 with Sasol

Gas on a willing buyer willing seller basis. NERSA published a consultation document

setting out its preliminary determination on the tariff application on 6 February 2012 in which

it proposed a substantial tariff reduction to the GTA2 tariff. On 21 February NERSA

informed ROMPCO that it should amend its application to conform to one of the four

methodologies set out in the NERSA Guidelines document. The present application was

prepared in response to this instruction from NERSA.

3 Methodology

9. The considerations for the tariff methodology adopted in this application are set out below.

3.1 Methodology selection: Discounted cash flow (Levelised cost)

10. The primary guidance utilised for the preparation of this tariff application was the NERSA

Guidelines document which lists a number of tariff methodology options from which

licensees can choose.

11. The discounted cash flow (“DCF”, also referred to herein as the levelised cost methodology)

was used for the calculation of the tariff in this application for the following reasons2:

1) The construction of the compressor station, which enabled the additional GTA2

volumes, was completed in 2010. The ROMPCO shareholders committed to this

investment on the basis of a discounted cash flow analysis of the project costs and

revenues generated by the GTA2 contractual tariffs. The discounted cash flow

2 A project discounted cash flow (DCF) model calculates the NPV of the project costs. When this cost is

converted to a cost per unit of service (in this case R/GJ transported) it is known as a “levelised cost”. In other words, as is demonstrated in this application, to calculate a tariff the NPV of costs have to be converted to a cost per unit of volume over the project life in such a way that the tariff applied to the volumes over the project life will generate revenues with an NPV equal to the NPV of project costs. We will thus use the term “levelised cost” to describe this (NPV based) method of calculating tariffs.

4

methodology thus best matches the basis on which the investment decision was

reached.

2) Levelised cost tariffs provide compensation later in the project life for the lower

volumes occurring during the initial gradual volume ramp-up stage of the project.

The levelised cost based tariff methodology spreads the lifetime revenue requirement

evenly over the project lifetime volumes and thus compensates for lower ramp up

volumes later in the life of the project. GTA2 has already progressed through two

years of the ramp up stage and thus now needs to be compensated for the lower

volumes during this stage. Switching to another tariff methodology (such as the rate-

of-return approach) would mean that shareholders are deprived of the compensation

for the loss of early stage project returns due to volume ramp up. The levelised cost

methodology, which is consistent with the NERSA Guidelines, ensures that this

problem does not occur.3

3) Further benefits of the levelised cost methodology includes that it is relatively straight

forward to implement (it focuses only on actual project cash flows, and does not

require the maintenance of an asset register, or complicated calculations related to

debt-pass through, etc.)

12. It should be recognised that using the prescribed NERSA DCF tariff methodology does not

compensate the shareholders to the extent that they expected when the investment was

made in the compressor on the basis of a willing buyer willing seller transportation

agreement and tariff. The internal rate of return (IRR) methodology used by the

shareholders, supported by banks providing loans, and which is no different from large

multinational companies when making investment decisions, sets out to achieve a minimum

IRR on an after tax and ungeared basis. The IRR is that discount rate, when applied to the

cash flows, that produces a zero net present value (NPV). The GTA2 tariff was set to

achieve this required return. The Capital Asset Pricing Method (CAPM) specification in the

NERSA Guidelines document does not allow for the same IRR to be achieved.

13. Given that one of the aims of regulation is to create an enabling environment for investment

in infrastructure, it is submitted that willing buyer willing seller contracts should be sustained.

This is, for instance, the case in the USA where such agreements are unregulated, are

viewed in a positive light and considered an indication of the competitiveness of the industry.

3 The Rate of Return methodology effectively applies a “snap-shot” view to calculate the tariff and therefore will

not compensate for lower tariffs during the ramp up stage as a result of the application of another methodology (the levelised cost methodology in this case).

5

3.2 Levelised cost implementation: the details

14. This section explains the details of the levelised cost method. As explained in footnote 2

above, levelised cost-based tariffs are calculated so that the tariff applied to the volumes

over the project life will generate revenues with an NPV equal to the NPV of the project

costs. A levelised cost tariff can either be fixed in nominal terms over the project life (leading

to very high tariffs early in the project life), or fixed in real terms (requiring annual indexation

by the inflation rate). The normal practice is to calculate tariffs that are fixed in real terms,

and this approach was also adopted for this application. The tariff is calculated according to

the following basic formula:

𝑇0 =𝑁𝑃𝑉 𝑜𝑓 𝑎𝑙𝑙 𝑝𝑟𝑜𝑗𝑒𝑐𝑡 𝑐𝑜𝑠𝑡𝑠

𝑆𝑢𝑚 𝑜𝑓 𝑑𝑖𝑠𝑐𝑜𝑢𝑛𝑡𝑒𝑑 𝑣𝑜𝑙𝑢𝑚𝑒𝑠

15. A comprehensive statement of the calculation is as follows:

𝑇0 =

𝐶𝑡 + 𝑂𝑀𝑡 + 𝐼𝑡 + 𝑡𝑡 (1 + 𝑟𝑛𝑚 )𝑡𝑚=1

𝑛𝑡=1

𝑉𝑡

(1 + 𝑟𝑟𝑚 )𝑡𝑚=1

𝑛𝑡=1

Where:

T0 : Fixed real tariff in period zero (the real cost based tariff is expressed in the

currency of period zero). T0 value will have to be indexed by the inflation rate to

calculate the tariff to charge in period t. 4

Ct : Investment cash flows in period t (nominal)

OMt : Operating and maintenance cash flow in period t (nominal)

It : All debt related cash flows (drawdown, repayment or interest) in period t

(nominal)

Tt : Tax payment cash flows in period t (nominal)

Vt : Volume in period t

rnm : Discount rate for period m (nominal)

rrm : Discount rate for period m (real)

4 If the real NPV of the costs are expressed in the currency value of period zero it will be equal to the nominal

NPV in period zero. The formula thus calculates the NPV of costs (the part above the line) in nominal terms to save the effort of having to convert all the cash flows to real values first.

6

16. Further aspects of the methodology adopted in the application are:

1) Discounting of cash flows and volumes is done one year at a time by using the

discount rate calculated for the year in question. This method allows for discount

rates to vary from year to year because of gearing changes5.

2) If 𝑇0 is to be expressed in real terms the volumes must be discounted by the real

discount rates, (and if it is to be expressed in nominal terms the volumes must be

discounted by the nominal discount rates).

3) As is the case for any discounted cash flow valuation, the calculation is entirely

based on cash flows and does not contain non-cash flow items such as depreciation,

etc.

4) In accordance with the debt pass through option set out in the NERSA Guidelines

Document the calculation included all financing related cash flows in the model. In

other words, debt draw downs, interest payments, and debt repayments are included

as cash flows. This has the benefit of accurately reflecting the actual level, timing

and cost of debt finance used for the project.

5) Because the actual debt related cash flows are included, the tax cost calculation can

also be more accurate by basing it on the expected after interest taxable profits for

each year.

6) Given that debt financing cash flows are included, it is furthermore necessary to

discount the cash flows (and volumes) at the cost of equity.

4 Selection of the system cost basis

17. ROMPCO operates in Mozambique and South Africa and transports gas to customers in

both countries. Furthermore, the present tariff application is only for a part of its sales in

South Africa (the GTA2 agreement). Given that NERSA has regulatory jurisdiction in South

Africa, the question about the correct system cost, ring fencing basis for calculating the

GTA2 tariffs for regulatory approval arises. This question is further complicated by the fact

that the costs of the pre-existing pipeline network are already being recovered by the GTA1

agreement, over which NERSA does not have tariff approval jurisdiction.

18. Two main system selection options were investigated for the purposes of determining the

GTA2 tariff, namely the “Compressor” option, and the “RSA Network” option. In the

Compressor option, only the incremental costs incurred to construct the compressor station

5 As is shown below, the tariff was finally modelled with a fixed gearing level.

7

at Komatipoort, which enabled the additional GTA2 volumes to be provided, is considered.

The “RSA Network” option considers only costs incurred (including assets constructed) in

South Africa. It apportions 55% of the total ROMPCO pipeline network costs plus the full

compressor station costs, and 55% of GTA1 volumes plus the GTA2 volumes for the tariff

calculation.

19. The discussion that follows sets out the framework for approaching the cost basis question,

before turning to the two approaches to evaluate their respective merits.

4.1 Framework for selecting the system cost basis

20. In the context of the question outlined above, this section develops the framework for

identifying the costs that should form the basis for the GTA2 tariff. The key elements of the

rationale comprising this framework are as follows:

NERSA’s jurisdiction

21. NERSA has jurisdiction for tariffs charged for service provided to customers in South Africa,

while it does not have approval jurisdiction over GTA1 tariffs.

Cost reflectivity

22. Tariffs should at least recover all costs (including the cost of capital) for providing service to

SA customers. While the Act is silent on the method to follow for determining transmission

tariffs, the Regulations do establish the principle that prices should reflect costs.

Furthermore, this principle is consistently established for the other sectors under NERSA‟s

jurisdiction. It can furthermore be concluded with a high level of confidence that most of the

objects of the Gas Act in effect require that the principle of cost reflectivity in the approval of

tariffs be adhered to.6 Lastly, a review of the regulatory principles outlined in the NERSA

Guidelines Document confirms that NERSA accepts the principle of cost reflectivity.

Cost allocation and revenue attribution

23. A fundamental implication of the principle of cost reflectivity is the requirement for cost and

revenue attribution. In other words, costs must be associated with, or attributed to, the

service provided. Tariffs should be designed so that the revenues generated by a tariff are

attributed to the specific costs incurred to provide the service for which the tariff is being

charged. The relationship can be illustrated as follows:

6 The objects in question are those listed in Section 2 of the Act, in particular: (a) – (e); and (h) – (j).

8

Causes causes (via tariffs)

SERVICE PROVIDED ▬► COSTS ▬► REVENUES

24. Furthermore, cost reflectivity requires that the relationship between costs and revenues must

be a one-to-one relationship: costs cannot be recovered twice, by more than one source of

revenue (say two tariffs under two different supply agreements), and all costs associated

with the provision of a particular service must be recovered by the revenue generated by the

tariff for that service. If the one-to-one relationship does not hold, the principle of cost

reflectivity will be breached.

25. The principle of attributing tariffs and associated revenues to the costs incurred to provide

the service has particular implications in the case of ROMPCO: while NERSA‟s jurisdiction is

limited to approving tariffs for service to South African customers only, all costs incurred by

ROMPCO to provide the service should be included in the tariff calculation – irrespective of

where the costs are incurred. This implies that the country where assets are located is not

relevant; rather what is relevant, for instance, is whether specific capital expenditure was

incurred to provide the service for which the tariff is being set.

26. At least three further interrelated questions often have to be resolved to determine the

appropriate cost allocation basis in network industries:

a) the degree to which the costs of different parts of the network are pooled7;

b) the degree to which tariffs structures and charges will be based on incremental costs

versus average costs8; and

c) the degree to which cross-subsidisation will be implemented between customers.

27. The implications of the framework set out here and the impact of these three questions for

the present GTA2 tariff application are discussed below

4.2 The implications for the GTA2 tariff

28. While the three cost allocation questions are important for a network operator and regulator

to resolve, they mostly do not arise with the present application for GTA2 tariffs. The

reasoning is as follows:

7 In South Africa this debate is at times characterised as the “Causer pays approach” versus the “System based

approach” for tariffs.

8 In microeconomic theory the term “marginal cost” is generally used rather than “incremental cost”. “Marginal

cost” refers to the cost of providing one additional unit of service (e.g. transporting one additional unit of gas volume), while incremental cost refers to the additional cost implications of a particular managerial action (for example an investment decision to increase capacity). For tariff calculation purposes it is generally more practical to work with “incremental costs”.

9

29. Question (a):

30. With the GTA1 agreement in place, the largest part of the ROMPCO cost “system” is not

currently under consideration. In other words, even if pooling costs to follow a system-based

approach was desirable, it is not currently possible with the legislative protection for GTA1.

There are two reasons for this conclusion. Firstly, the requirement for a one-to-one

relationship between costs and revenues will be breached if the entire system costs (or an

RSA apportionment thereof) is pooled and proportionally allocated to GTA2. (This will have

the effect of increasing the GTA1 cost base). In other words, the one-to-one relationship

between costs and revenues (tariffs) requires that the cost allocated to the GTA1 tariffs

should not be changed. Secondly, the purpose of the legal protection of the GTA1 tariff was

to ensure the return that shareholders received on the MSP would remain unchanged, while

ensuring an acceptable tariff to the South African market. Shifting compressor related costs

onto the GTA1 cost base by pooling costs for the two tariffs as part of the GTA2 calculation

will also breach the GTA1 legal protection.

31. In other words, cost pooling for the purposes of the GTA2 calculation is not possible

because the GTA1 tariff and margin cannot be changed. This conclusion suggests that the

RSA network cost allocation basis will not be a viable basis for setting GTA2 tariffs.

32. Question (b):

33. Because cost pooling is not possible across GTA1 and GTA2, the cost base for GTA2 will

necessarily have to reflect the incremental cost incurred to enable this additional service to

be provided in South Africa. This conclusion points to the use of the “Compressor” option

approach as a basis for GTA2 cost allocation. For the GTA2 evaluation it is thus not

necessary or possible for NERSA to adopt an in-principle position on the question of

average vs. incremental (or marginal) costing.

34. Question (c):

35. Currently ROMPCO operates a single pipeline with one off-take point and one customer in

Mozambique, and one off-take point and one customer in South Africa. Furthermore,

presently NERSA only evaluates tariffs for non-GTA1 sales in South Africa (currently only

GTA2). The question of cross subsidisation therefore does not arise, because no cross-

subsidisation is possible.

36. Taken together, the framework outlined above, and the assessment of the applicability of the

three cost allocation questions for the case of GTA2, leads consistently to the conclusion

that the “Compressor” option is the only cost basis for GTA2 that adheres to the principles of

cost reflectivity and the requirements of legality.

10

4.3 An alternative view

37. Given the conclusion above, it is worthwhile considering the alternative option: the “RSA

Network” option, as a cost basis for the GTA2 tariffs. A number of difficulties with this

approach arise.

38. It is difficult, if not impossible, to achieve the allocation of ROMPCO assets in a way that

meaningfully complies with the requirement to be “in South Africa”. For instance it has not

been recorded what the cost of the portion of the pipeline is that is in South Africa. Currently

the identification of South African assets and costs is emulated based on estimations of

costs and effort to build the pipeline, by apportioning 55% of the total ROMPCO assets,

costs (including tax) and volumes to South Africa. Dividing 55% of the costs by 55% of the

volumes in effect amounts to evaluating the entire cost base and volume base – it is not

possible to achieve an actual split or allocation of the assets.

39. Furthermore, even if a true asset and cost allocation was possible, The RSA Network

rationale does not adhere to the requirements of cost reflectivity and legality as outlined

above. This conclusion can be demonstrated as follows: Tariffs could be set for the full

GTA2 service (from Temane to Secunda) as follows:

Tariff = 55% x total ROMPCO costs / 55% x total ROMPCO volumes (i.e. using a 55%

apportionment)

= total ROMPCO costs / total ROMPCO volumes

40. In this case the tariff will not be based on the cost of providing GTA2 service to South African

customers alone. It will consider costs recovered by GTA1 over which NERSA does not

have jurisdiction.9

4.4 System cost basis: conclusions

41. The conclusions from the discussion above are the following:

42. The cost reflectivity and revenue attribution principles mean that all costs for providing a

service, irrespective of location, should be recovered in the tariff charged therefore.

9 Still assuming that true cost allocation is possible (it has already been argued above that it is not), another cost

basis option would be to set the GTA2 tariff for the South African part of the transport service only (Komatipoort - Secunda). ROMPCO will thus have to charge South African customers twice (Temane - Komatipoort, Komatipoort – Secunda). This option will breach the current GTA2 supply agreement, which is a full service contract (Temane - Secunda), and thereby significantly increase regulatory risk and undermine investor confidence in the sector. Investors will also face a further increase in the regulatory complexity of cross border pipelines. South African customers will not have regulatory protection for the Mozambique portion of the transportation: It is unclear whether the Mozambique regulator will be able to regulate tariffs up to the border for customers in South Africa. It appears that regulatory complexity and risk would be significantly reduced if the South African regulator sets tariffs for full service to South African customers, and Mozambique regulator sets tariffs for service to Mozambique customers.

11

43. If the “RSA network” options were to be used as a cost allocation approach, the trunk line

assets that are physically located in South Africa will have to be identified. However,

information on the physical location and cost of the specific pipeline component is not

currently available; these assets can only be allocated on the basis of apportioning the total

pipeline asset value between the two countries. In effect this means that the “RSA network”

option basis has to consider the costs of the entire pipeline.

44. GTA1, however, creates an unusual regulatory situation that does not allow the option of

implementing system based tariffs (as required by the “RSA network” option). Existing

pipeline costs are already being recovered under the GTA1 agreement and cannot be

considered for GTA2 tariffs, and neither can part of the GTA2 costs be shifted onto the

GTA1 revenue base by means of cost pooling.

45. The only remaining cost allocation basis for GTA2 is to adopt an incremental cost approach,

i.e. the “Compressor” option. The “Compressor” option has the further benefit that it

accurately reflects the basis on which the decision to invest in this system capacity

expansion was made: the additional capital expenditure on the compressor station was

justified on the basis of the additional revenues that could be generated from the additional

volumes. Regulatory decisions that support the basis for private sector investment in the

piped gas industry will be consistent with the objects of the Act.

46. Lastly, identifying the incremental costs of the “Compressor” option is straight forward and

uncontroversial.

47. Taken together the facts and arguments presented here thus consistently point to the

conclusion that the “Compressor” option is the only viable approach for determining the cost

basis for the GTA2 tariffs.

5 The cash flows

48. The detailed cash flows used in the levelised cost calculation are described below. All

amounts are shown in nominal terms and grouped by the ROMPCO financial year cycle

which ends on 30 June each year.10

10 Cash flows are deemed to occur at the end of each period. The compressor station was commissioned during

FY2011. The project start date is modelled as the end of the previous FY, which is at 30 June 2010. Capital expenditure mostly occurred before this date and is thus expressed in the model as a future value at 30 June 2010.

12

5.1 Capital expenditure

49. In accordance with the Compressor cost basis adopted for the tariff calculation, the capital

expenditure cash flows modelled consists only of the costs related to the decision to

construct the compressor station at Komatipoort. The details of the expenditure are shown

in the table below.

Table 1: Capital expenditure

Capital Expenditure FY 08 FY 09 FY 10 FY 11 FY 12 FY 13 Total

Servitude R'm 2.2

2.2

Factory Buildings R'm

8.2

8.2

Rest of Compressor R'm

894.3 10.6

904.9

Computer Equipment R'm

3.5

3.5

Vehicles R'm 0.8

2.1

2.9

Interest Capitalised R'm

18.9

18.9

Black Powder R'm

28.0 28.0

Total R'm 3.0 - - 926.9 10.6 28.0 968.5

50. The cash flows also include an allowance for decommissioning expenditures of R30.3m

occurring in FY2029 (not shown in the table above).

5.2 Operating costs

51. The Government of the Republic of South Africa and Sasol Limited concluded an

“Agreement concerning the Mozambican Gas Pipeline” (“Regulatory Agreement”) dated 26

September 2001. Clause 10 of the Regulatory Agreement provides for the appointment of

Sasol Gas as the Operator of the MSP as well as the basis for the operating fee, namely

cost plus 10% (clause10.5). Clause 10 of the agreement also provides that annual audits

will be conducted to verify the cost competitiveness of the Sasol‟s operation of the pipeline.

The table below shows the operating costs used in the levelised cost calculation.

Table 2: Total operating costs

Period Variable Costs

Fixed Costs

Decomm Costs

Total

R'm R'm R'm R'm

FY 10 - - - -

FY 11 6.8 39.9 - 46.8

FY 12 8.7 42.3 - 51.0

FY 13 10.4 44.6 - 55.0

FY 14 13.7 47.4 - 61.1

FY 15 18.3 50.2 - 68.5

FY 16 19.9 53.0 - 72.9

FY 17 21.0 55.7 - 76.7

FY 18 22.2 58.5 - 80.7

13

Period Variable Costs

Fixed Costs

Decomm Costs

Total

R'm R'm R'm R'm

FY 19 23.2 61.5 - 84.6

FY 20 24.4 64.6 - 89.0

FY 21 25.7 67.9 - 93.6

FY 22 26.8 71.3 - 98.2

FY 23 28.2 74.9 - 103.1

FY 24 29.6 78.7 - 108.3

FY 25 31.1 82.7 - 113.8

FY 26 32.7 86.9 - 119.6

FY 27 34.4 91.3 - 125.6

FY 28 36.1 95.9 - 132.0

FY 29 37.9 100.8 30.3 169.0

451.3 1,268.2 30.3 1,749.7

52. Variable costs consist entirely of burner and compressor gas. Further details on the fixed

costs are shown in Appendix B.

5.3 Debt finance

53. The details of the loan agreements relating to the Compressor station are as follows:

The initial revolving bridging finance facility of R1,1 billion was replaced with a R600

million medium term finance facility from CEF (Pty) Limited (“CEF”) and Sasol

Financing (Pty) Limited (“Sasol Financing”).

During November 2009 the Board approved that ROMPCO enters into two separate

loan agreements with Deutsche Investitions – und Entwicklungsgesellschaft mbH

(“DEG”) and with The Standard Bank of South Africa Limited (“SBSA”). ROMPCO

approached its shareholders to provide the necessary guarantees.

During May 2010 the Board approved a 92% guaranteed agreement with DEG for

R250 million and with SBSA for R350 million independently, both with a term of 3.5

years.

ROMPCO‟s debt: equity ratio decreased from 0,65 in FY11 to 0,43 in FY12.

54. For the purposes of the levelised cost calculation it is necessary to incorporate gearing

information for the entire project duration. However, existing financing plans do not cover

this time frame. Rather than reflecting the information set out above (which implies 0%

gearing after 2014), the approach adopted was thus that ROMPCO‟s policy of maintaining a

target debt level over the project life should be reflected. This was achieved by means of

replacing the actual loans in the model with a “synthetic” loan that shows annual interest

payments and adjustments in the debt level to maintain the gearing target throughout the

project life. ROMPCO‟s target gearing is 45%, and while it aims to maintain this gearing

14

level, practical considerations mean that gearing could, on the short-term, vary around the

target level in a band between 40% - 60%.

55. Based on the cost of debt for the recent loans associated with the compressor station, as

shown in the table below, the cost of debt for the synthetic loan was estimated at 3 month

JIBAR plus 2.5%.

Table 3: Cost of debt

Cost of Debt Estimation

Details SBSA DEG SYNTHETIC LOAN

Debt Instrument Term Loan Term Loan Revolving Loan

Tenor 3.5 3.5 19.0

Reference Rate 3 Month Jibar 6 Month Jibar 3 Month Jibar

Margin 1.68% 3.40% 2.50%

Start 20-Jun-11 20-Jun-11 01-Jul-10

End 15-Dec-14 15-Dec-14 30-Jun-29

56. The details of the debt related cash flows are shown in Appendix B.

5.4 Tax

57. ROMPCO, being a South African resident, is taxed on its total world-wide taxable income

(i.e. including the income of its Mozambican branch) and rebates are granted against the

South African tax payable (double taxation agreement “DTA”) in respect of the foreign taxes

on such income up to the amount of the relevant SA taxes, provided the income is from a

non-SA source.

58. ROMPCO followed the pass-through tax calculation approach for the levelised cost

calculation, and thus modeled the tax obligation that arises in each year as the relevant tax

cash flow. The asset classification shown in Appendix B was used as the basis to calculate

the capital, and wear and tear allowances.

6 Discount rate

59. The cash flows used in the levelised cost calculation includes the actual debt related cash

flows. The discount rate to be used should thus not reflect the cost of debt, as this has

already been taken into account.11 The appropriate discount rate will then be the cost of

equity. Because cash flows were modelled in nominal terms (after, or including inflation) the

11 It would thus not be appropriate to use the weighted average cost of capital (WACC) for discounting purposes.

15

discount rate should be expressed in nominal terms as well. The formula used for the cost

of equity is as follows:

Ke = β * MRP + Rf + (SSP + CRP)

Where:

Β : Beta, using system gearing, βasset = 0.642

MRP : Market Risk Premium, 5.5% (real)

Rf : Risk free rate, 4.4% (real)

SSP : Small Stock Premium, 3.2% (real)

CRP : Second country risk premium, 3.4% (real)

Ke : Cost of equity, 16.65%(real), 23.13%(nominal) @ 45% gearing12

60. The details of the cost of equity calculation for every year are shown in the tables in

Appendix B. The elements of the formula are discussed below.

6.1 Beta

61. The equity beta measures the systematic (non-diversifiable) risk of the activity in question.

In order to satisfy the requirements of NERSA‟s Guidelines Document, ROMPCO needed to

obtain equity betas for six suitable listed proxy companies to estimate its own equity beta.

62. In its consultation document regarding the ROMPCO pipeline‟s preliminary gas transmission

tariff determination for 2011/12, NERSA has used the following six US companies as

proxies:

AGL Resources Inc.

UGI Corporation

South Jersey Industries

WGL Holdings Inc.

The Laclede Group

Piedmont Natural Gas Company Inc.

63. ROMPCO has therefore used these six companies as the proxies in the estimation of the

equity beta for ROMPCO.

12 This cost of equity compares well with NERSA‟s decision to award Petroline RSA PTY(Ltd) a cost of equity of

16.82% (real) on 18 June 2008.

16

64. The tariff methodology in NERSA‟s Guidelines Document sets out the required procedure for

estimating the equity beta of regulated gas transmission companies. The guidelines are

quite prescriptive in some areas. For example, NERSA mandates that:

six quoted international companies should be used as proxy firms when estimating

the beta of the regulated business in South Africa;

the asset betas of these six proxy companies should be size-weighted and averaged

in order to obtain the asset beta of the regulated company in South Africa;

the Hamada formula must be used for:

o unlevering the estimated equity betas of the proxy companies to obtain their

asset betas; and

o relevering the average asset beta calculated for the regulated company to

obtain the final equity beta for ROMPCO.

65. However, the guidelines are silent on a number of important methodological choices,

specifically in the estimation and adjustment of the equity betas of the proxy companies.

These choices are related to the following issues:

The frequency of returns data used in the estimation;

The appropriate estimation window; and

The case for adjustments to raw betas.

66. In relation to this, ROMPCO has made the following choices in its estimation of the equity

betas of the proxy companies:

daily returns observations;

an estimation window using returns measured over the two-year period to 23 March

2012; and

the Vasicek approach to adjust the estimated raw equity betas.

67. ROMPCO views these choices as reasonable from a methodological standpoint. Using the

estimated equity betas for the proxy companies, we have used the Hamada formula to

calculate their corresponding asset betas.

68. The last step in the determination of the equity beta for ROMPCO has been to calculate its

asset beta and then relever it to obtain its final equity beta. Following NERSA‟s guidelines,

the asset beta for ROMPCO was calculated as the weighted average of the estimated asset

betas of the proxy companies. The relative size of each proxy company was used,

measured as the sum of the average market capitalisation and average debt, as the weights

applied in this average calculation. Finally, the asset beta has been relevered, using the

17

Hamada formula for each year in question based on the gearing applicable at that time, to

obtain the equity beta for ROMPCO13.

6.2 Risk Free Rate and Market Risk Premium

69. NERSA has developed a spreadsheet for the calculation of the risk free rate and market risk

premium to use for its tariff determinations for the Petroleum Pipelines sector. ROMPCO

obtained a copy of the spreadsheet from NERSA‟s website, and updated the data therein

(using the same sources) up to December 2011. The data analysed covers a period of 25

years over monthly intervals. The updated NERSA spreadsheet is submitted with this

application as Appendix C.

6.3 Small Stock Premium

70. It is common practice to add a premium to the cost of equity for smaller companies.

Damodaran reports that analysts in the US typically add a premium of between 3 – 3.5% for

smaller companies14. PWC South Africa conducts regular valuation surveys, and report the

following results for the use of SSP in valuations from their 2009/10 survey:

Figure 2: Small Stock Premium applied in South Africa

71. It is noteworthy that since March 2011 NERSA has also recognised the use of a small stock

premium in its tariff methodology for the Petroleum Pipelines Industry, and has previously

awarded small stock premiums to Transnet Limited and Petroline RSA PTY(Ltd).

72. The value used in the ROMPCO calculation was derived from the table above for companies

with a turn-over between Rm 251 - Rm500, converted to a value of 3.2% (real).

13 Appendix One contains the detailed report on the calculation of the appropriate equity beta for ROMPCO,

commissioned from Frontier Economics.

14 Damodaran, A. 2009. Advanced Valuation. Seminar presentation. p.44. Accessed at

http://www.damodaran.com.

18

6.4 Second Country Risk Premium

73. The viability of all the phases of the ROMPCO project is exposed to the country risks of both

South Africa and Mozambique. The risk free rate utilised in the cost of equity calculation

already reflects the country risk rate of South Africa. A separate parameter has to be

included to reflect the country risk rate for Mozambique. The analysis presented here

follows the methodology of Damodaran (2012)15 at the New York University, Stern School of

Business. The calculations are shown in the table below:

Table 4: Second country risk adjustment

Sovereign risk rating

Adjusted spread over US risk free rate

Equity risk premium

Mozambique B1 400 6.00%

RSA A3 115 1.73%

Difference 4.28%

Beta @ 30% gearing (lowest gearing)

84.10%

Country risk adjustment 3.59%

74. Overall it has to be noted that the calculated discount rate presented above, using NERSA‟s

method, is lower than shareholder expectations at the time when the investment was

undertaken and when the ship-or-pay agreement was established on a willing buyer – willing

seller basis. Such large changes in project value may affect future investment decisions in

the sector.

75. For instance the following risk factors are not adequately incorporated in the NERSA

Methdology: specific company risk, including: reservoir risk (which can negatively affect the

cash flow period over which to recoup investment); regulatory uncertainty due to the

investment spanning two countries; the impact of regulation on the industry, which could

negatively affect demand for transmission services.

7 Volume forecasts

76. To date the GTA2 volumes have been below the target levels and the ship or pay

mechanism has been triggered every year since its operation. ROMPCO‟s budgeting is

15 See http://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/ctryprem.html, accessed in July 2012.

19

therefore based on the ship or pay volumes, and this was also the basis for the tariff

calculation presented here. The detailed volume forecasts are shown in Appendix B.

8 Tariff calculation

77. The details of the levelised cost calculation are shown in the table below.

Table 5: Levelised cost calculation

Levelised Cost Calculations

Asset Levelised

Cost

Financing Levelised

Cost

Opex & Working Capital Levelised

Cost Total

NPV (FY10) 797.1 (164.8) 361.0 993.4

Discounted Volumes 124.6 124.6 124.6 Levelised Cost (FY10) 6.397 (1.322) 2.897 7.972

Levelised Cost (FY11) 8.373

Levelised Cost (FY12) 8.866

Levelised Cost (FY13) 9.359

Test: NPV of (tariff in each year x volumes) 797.1 (164.8) 361.0

*Finance levelised cost is negative because the debt repayments and interest payments are discounted at the higher cost of equity.

78. The annual tariffs shown here and in Appendix B is the average tariff level for each year. In

practice, and in accordance with the GTA2 agreement, the tariff will be adjusted quarterly by

the rate of inflation as follows:

𝑇𝑛 = 𝑇0

𝐶𝑃𝐼𝑛𝐶𝑃𝐼0

Where:

Tn : Tariff in quarter n. Quarterly periods are counted from period zero which is in

the middle of the last ROMPCO financial year ending before the first regulatory

tariff approval (FY ending 30 June 2012).

T0 : Reference tariff (1 January 2012) = average tariff calculated for FY2012

(R8.866/GJ) adjusted to the level for 1 January 2012.16

𝑇0 = 𝑇average

(𝐶𝑃𝐼0 − 𝐶𝑃𝐼−1)2 + 𝐶𝑃𝐼−1

𝐶𝑃𝐼−1

16 The equivalent of half a period adjustment is required because the tariff for period zero is below the average

tariff for the year and the tariff for period one is above the average tariff.

20

CPIn : The arithmetic average of the South African Consumer Price Index for all

commodities for consumption in South Africa as reported by Statistics South

Africa for the three months preceding the last day of the first month of the

Quarter immediately preceding the price adjustment date for which the price

calculation is done;

CPI-1 , CPI0: CPI index for period minus one and period zero (31 October 2011 and 31

December 2011) calculated on the same basis as for CPIn

79. The calculation of future tariffs presented to NERSA for approval is thus as follows:

An average tariff level for FY2012 of R8.866/GJ, for the purposes of calculating the

reference tariff (T0) as set out above;

Quarterly indexation by CPI as set out above;

A quarterly price adjustment at 12h00 am on the first day of each quarter, for the

duration of the project life.

9 Conclusions

80. The Komatipoort compressor station was constructed to enable the provision of the

additional volumes supplied under GTA2. The investment was undertaken by the

shareholders on the basis of the GTA2 tariff specified in the term sheet concluded between

ROMPCO and Sasol Gas. The revision to these tariffs as envisaged by NERSA will

substantially alter the expected returns on the project.

81. In accordance with the instruction from NERSA the tariff calculation presented in this

document follows the discounted cash flow / levelised cost method as described in the

NERSA Guidelines document.

82. Regulatory approval of the tariff and its adjustment structure for the duration of the project

life as proposed in this document will contribute to reducing regulatory risk and encourage

further private sector investment in the piped gas transmission sector.

21

Appendix A: Determination of the Equity Beta for ROMPCO

The report by Frontier Economics is provided separately.

22

Appendix B: Detailed tables

Table 6: Fixed operating costs

Period Labour Maintenance

Material Rentals

Sundry Costs

Pipeline management

fees Total

R'm R'm R'm R'm R'm R'm

FY 10 - - - - - -

FY 11 10.2 21.9 0.8 3.3 3.7 39.9

FY 12 10.8 23.2 0.8 3.5 4.0 42.3

FY 13 11.4 24.5 0.9 3.7 4.2 44.6

FY 14 12.1 26.0 0.9 4.0 4.4 47.4

FY 15 12.8 27.5 1.0 4.2 4.7 50.2

FY 16 13.5 29.1 1.0 4.4 5.0 53.0

FY 17 14.2 30.5 1.1 4.7 5.2 55.7

FY 18 14.9 32.1 1.1 4.9 5.5 58.5

FY 19 15.7 33.7 1.2 5.1 5.7 61.5

FY 20 16.5 35.4 1.3 5.4 6.0 64.6

FY 21 17.3 37.2 1.3 5.7 6.3 67.9

FY 22 18.2 39.1 1.4 6.0 6.7 71.3

FY 23 19.1 41.1 1.5 6.3 7.0 74.9

FY 24 20.1 43.1 1.5 6.6 7.4 78.7

FY 25 21.1 45.3 1.6 6.9 7.7 82.7

FY 26 22.2 47.6 1.7 7.3 8.1 86.9

FY 27 23.3 50.0 1.8 7.6 8.5 91.3

FY 28 24.5 52.6 1.9 8.0 9.0 95.9

FY 29 25.7 55.2 2.0 8.4 9.4 100.8

323.8 694.9 24.9 106.0 118.5 1,268.2

Table 7: Components of the cost of equity calculation

Components of Cost of Equity

Risk Free Rate % Real 4.4% Market Risk Premium % Real 5.5% Asset Beta

Real 0.64

2nd County Risk Adjustment % Real 3.4% Small Stocks Premium % Real 3.2%

23

Table 8: Details of cost of equity calculation (1)

Period Debt Portion of

Capital Structure Debt/Equity Equity Beta

FY 10 0% - 0.64

FY 11 0% - 0.64

FY 12 45% 0.82 1.02

FY 13 45% 0.82 1.02

FY 14 45% 0.82 1.02

FY 15 45% 0.82 1.02

FY 16 45% 0.82 1.02

FY 17 45% 0.82 1.02

FY 18 45% 0.82 1.02

FY 19 45% 0.82 1.02

FY 20 45% 0.82 1.02

FY 21 45% 0.82 1.02

FY 22 45% 0.82 1.02

FY 23 45% 0.82 1.02

FY 24 45% 0.82 1.02

FY 25 45% 0.82 1.02

FY 26 45% 0.82 1.02

FY 27 45% 0.82 1.02

FY 28 45% 0.82 1.02

FY 29 45% 0.82 1.02

Table 9: Details of cost of equity calculation (2)

Period

Market Risk Premium x Equity Beta

Risk Free Rate

2nd County Risk Adjustment

Small Stocks Premium

Cost of Equity (Ke)

Inflation Cost of Equity (Ke)

Nominal Discount Factors

Nominal Discount Factors

% % % % Real CPI Nominal 2010 2013

FY 10 3.5% 4.4% 3.4% 3.2% 14.6% 4.206% 19.4% 1.000 -

FY 11 3.5% 4.4% 3.4% 3.2% 14.6% 5.022% 20.3% 0.831 -

FY 12 5.6% 4.4% 3.4% 3.2% 16.6% 5.892% 23.5% 0.673 -

FY 13 5.6% 4.4% 3.4% 3.2% 16.6% 5.558% 23.1% 0.546 1.000

FY 14 5.6% 4.4% 3.4% 3.2% 16.6% 6.122% 23.8% 0.441 0.808

FY 15 5.6% 4.4% 3.4% 3.2% 16.6% 5.885% 23.5% 0.357 0.654

FY 16 5.6% 4.4% 3.4% 3.2% 16.6% 5.683% 23.3% 0.290 0.531

FY 17 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.237 0.433

FY 18 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.193 0.353

FY 19 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.158 0.288

FY 20 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.129 0.235

FY 21 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.105 0.192

FY 22 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.086 0.157

FY 23 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.070 0.128

FY 24 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.057 0.104

FY 25 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.046 0.085

FY 26 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.038 0.069

FY 27 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.031 0.057

FY 28 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.025 0.046

FY 29 5.6% 4.4% 3.4% 3.2% 16.6% 5.063% 22.6% 0.021 0.038

24

Table 10: Debt related cash flows

Period Original Debt Change in

Debt Revised Debt

Interest Rate

Total Interest

Paid

R'm R'm R'm % R'm

FY 10 - - - 0.0% -

FY 11 600.0 -170.1 429.9 8.1% 48.5

FY 12 398.4 24.5 422.9 8.1% 32.3

FY 13 400.8 31.8 432.6 8.1% 32.5

FY 14 409.0 23.6 432.6 8.1% 33.1

FY 15 407.1 24.0 431.0 8.1% 33.0

FY 16 403.9 24.5 428.4 8.1% 32.7

FY 17 399.5 23.7 423.2 8.1% 32.4

FY 18 392.5 24.9 417.4 8.1% 31.8

FY 19 384.7 26.3 411.0 8.1% 31.2

FY 20 376.0 27.7 403.7 8.1% 30.5

FY 21 366.1 29.5 395.6 8.1% 29.7

FY 22 354.9 31.5 386.4 8.1% 28.7

FY 23 342.1 33.5 375.5 8.1% 27.7

FY 24 326.7 35.1 361.8 8.1% 26.5

FY 25 307.3 35.0 342.3 8.1% 24.9

FY 26 280.9 28.2 309.1 8.1% 22.8

FY 27 238.8 45.6 284.4 8.1% 19.3

FY 28 204.8 41.9 246.7 8.1% 16.6

FY 29 144.3 - 144.3 8.1% 11.7

6,737.7 341.2 7,078.9

545.6

25

Table 11: Historic and forecast volumes

Volumes Period MGJ/a

FY 10 -

FY 11 16.6

FY 12 21.2

FY 13 18.0

FY 14 19.8

FY 15 21.6

FY 16 23.5

FY 17 24.8

FY 18 24.8

FY 19 24.8

FY 20 24.8

FY 21 24.8

FY 22 24.8

FY 23 24.8

FY 24 24.8

FY 25 24.8

FY 26 24.8

FY 27 24.8

FY 28 24.8

FY 29 24.8

443.6

26

Table 12: Tax treatment of assets

Account Description

Accounting Treatment Tax Treatment

Useful Life

% Rate p.a

Useful Life

% Rate p.a

1500 Land and buildings 50 2% 20 5%

11026 workshp Equipment 10 10% 5 20%

11097 Pipelines > 28 Feb 2001 25 4% 10 10%

Rest of Compressors 19 5.3% 10 10%

Compressors Turbine Engine 15 6.7% 10 10%

13036 Commercial vehicles 5 20% 4 25%

14016 Computer Equip > R5000 3 33% 3 33,3%

25006 Computer equip < R5000 1

100% on acquisition 3 33,3%

25026 Depot Equip < R5000 1

100% on acquisition 5 20%

32006 Cap invest measure 0 None 0 None

34026 Computer software (PIMS) 3 33,3% 3 33,3%

34007 Servitudes 25 4% Not

deducted None

27

Table 13: Tax calculation

Tax Calculation

Period EBITDA Interest Wear & tear

Capital Allowances

Taxable Income/(loss) before loss Utilisation

Assessed Losses BEG

Assessed Losses END

Taxable Income after loss Utilisation

Company Tax Rate

Tax expense

Nominal R'm R'm R'm R'm R'm R'm R'm R'm R'm

FY10 - - -0.2 - -0.2 -0.3 -0.5 - 0.3 -

FY11 92.2 -48.5 -1.2 -91.8 -49.3 -0.5 -49.8 - 0.3 -

FY12 136.7 -32.3 -1.1 -92.9 10.5 -49.8 -39.4 - 0.3 -

FY13 113.2 -32.5 -1.1 -95.7 -16.0 -39.4 -55.4 - 0.3 -

FY14 135.9 -33.1 -0.8 -95.7 6.4 -55.4 -49.0 - 0.3 -

FY15 158.9 -33.0 -0.2 -95.7 30.0 -49.0 -19.0 - 0.3 -

FY16 187.8 -32.7 -0.2 -95.2 59.7 -19.0 - 40.7 0.3 11.4

FY17 213.3 -32.4 -0.2 -95.2 85.5 - - 85.5 0.3 23.9

FY18 224.0 -31.8 -0.2 -95.2 96.8 - - 96.8 0.3 27.1

FY19 235.5 -31.2 -0.2 -95.2 108.9 - - 108.9 0.3 30.5

FY20 247.3 -30.5 -0.2 -95.2 121.4 - - 121.4 0.3 34.0

FY21 259.8 -29.7 -0.2 -3.9 226.0 - - 226.0 0.3 63.3

FY22 273.1 -28.7 -0.2 -2.8 241.3 - - 241.3 0.3 67.6

FY23 287.0 -27.7 -0.2 - 259.0 - - 259.0 0.3 72.5

FY24 301.5 -26.5 -0.2 - 274.8 - - 274.8 0.3 76.9

FY25 316.8 -24.9 -0.2 - 291.6 - - 291.6 0.3 81.7

FY26 332.8 -22.8 -0.2 - 309.8 - - 309.8 0.3 86.7

FY27 349.7 -19.3 -0.2 - 330.1 - - 330.1 0.3 92.4

FY28 367.4 -16.6 -0.2 - 350.5 - - 350.5 0.3 98.1

FY29 355.6 -11.7 -0.2 - 343.7 - - 343.7 0.3 96.2

4,588.6 -545.6 -8.2 -954.2 3,080.6 -213.4 -213.1 3,080.3 5.6 862.5

28

Table 14: Discounted cash flow calculation

DCF

Period Average

Tariff Revenue

Variable Costs

Fixed Costs Decomm

Costs Tax Capex

Working Capital

Funding Capital

Repayment Interest

Free Cash flow to Equity

R/GJ R'm R'm R'm R'm R'm R'm R'm R'm R'm R'm R'm

FY10 7.972 - - - - - -3.9 - - - - -3.9

FY11 8.373 138.9 -6.8 -39.9 - - -926.9 -11.4 429.9 -31.6 -48.5 -496.2

FY12 8.866 187.8 -8.7 -42.3 - - -10.6 -5.4 24.5 -22.1 -32.3 90.8

FY13 9.359 168.3 -10.4 -44.6 - - -28.0 2.8 31.8 -23.6 -32.5 63.9

FY14 9.932 197.0 -13.7 -47.4 - - - -2.8 23.6 -25.6 -33.1 98.1

FY15 10.516 227.4 -18.3 -50.2 - - - -2.8 24.0 -27.1 -33.0 119.9

FY16 11.114 260.7 -19.9 -53.0 - -11.4 - -3.5 24.5 -28.8 -32.7 135.9

FY17 11.677 290.0 -21.0 -55.7 - -23.9 - -3.2 23.7 -30.7 -32.4 146.7

FY18 12.268 304.7 -22.2 -58.5 - -27.1 - -1.3 24.9 -32.7 -31.8 156.0

FY19 12.889 320.2 -23.2 -61.5 - -30.5 - -1.4 26.3 -35.0 -31.2 163.7

FY20 13.541 336.4 -24.4 -64.6 - -34.0 - -1.4 27.7 -37.6 -30.5 171.6

FY21 14.227 353.4 -25.7 -67.9 - -63.3 - -1.6 29.5 -40.7 -29.7 154.0

FY22 14.947 371.3 -26.8 -71.3 - -67.6 - -1.6 31.5 -44.4 -28.7 162.3

FY23 15.704 390.1 -28.2 -74.9 - -72.5 - -1.7 33.5 -48.9 -27.7 169.7

FY24 16.499 409.8 -29.6 -78.7 - -76.9 - -1.7 35.1 -54.4 -26.5 177.0

FY25 17.335 430.6 -31.1 -82.7 - -81.7 - -2.0 35.0 -61.5 -24.9 181.8

FY26 18.212 452.4 -32.7 -86.9 - -86.7 - -2.0 28.2 -70.2 -22.8 179.3

FY27 19.134 475.3 -34.4 -91.3 - -92.4 - -2.1 45.6 -79.6 -19.3 201.8

FY28 20.103 499.4 -36.1 -95.9 - -98.1 - -2.1 41.9 -102.4 -16.6 190.1

FY29 21.121 524.6 -37.9 -100.8 -30.3 -96.2 - 86.8 -0.0 -144.3 -11.7 190.3

6,338.3 -451.3 -1,268.2 -30.3 -862.5 -969.4 41.7 941.2 -941.2 -545.6 2,252.7

29

Appendix C: Updated NERSA risk free rate and market risk premium spreadsheet

Due to its size this spreadsheet will be provided in electronic format.