APPROVAL OF TARIFF METHODOLOGIES FOR USE IN THE PETROLEUM PIPELINES INDUSTRY. THE DECISION

ENERGY REGULATOR OF SOUTH AFRICA In the matter regarding APPROVAL OF TARIFF METHODOLOGIES FOR USE IN THE PETROLEUM PIPELINES INDUSTRY. THE DECISION On 31 March 2011 the Energy Regulator approved the following methodologies: i. Tariff for the Setting of Tariffs in the Petroleum Pipelines Industry: 5 th Edition Approved 31 March 2011; and ii. Tariff for the Approval of tariffs for Petroleum Facilities and Petroleum Storage Facilities: 2 nd Edition - Approved 31 March 2011 incorporating certain changes from previous editions. REASONS FOR THE DECISION Introduction 1. On 26 November 2009 the Energy Regulator approved the Tariff for the Setting of Tariffs in the Petroleum Pipelines Industry - 4 th Edition, and on 25 March 2010 the Energy Regulator approved the RfD: Tariff Methodologies: 2 nd Edition on Storage & tariffs and the 5 th edition on the setting of Pipeline tariffs: 31 March 2011. Page 1 of 5

Tariff for the Approval of Tariffs for Petroleum Facilities and Storage Facilities (the Methodologies). 2. The refinements to the Methodologies in this decision relate primarily to the formula for determining the cost of equity (Ke). Background to the changes proposed in the Tariff Methodologies 3. In the process of applying the tariff methodologies over the past few years, changes to the methodologies had been effected for the following reasons: - the promulgation of Regulations which were published in April 2008; - continual research and benchmarking by NERSA on developments within the field of economic regulation; and - comments received from stakeholders and licensed entities on the applicability of some of the principles and practises in these methodologies to their specific situations. 4. The most recent opportunity which has been utilised to reassess the appropriateness of the tariff methodologies, was the processing of Transnet Limited s 2011/12 petroleum pipeline tariff application. Rationale behind the Changes to the tariff methodologies 5. The proposed refinements to certain of the elements in the Methodologies have come about for the reasons provided below: - the Energy Regulator is endeavouring to move the tariff methodologies that it uses in the electricity, piped gas and petroleum pipelines industries towards increased convergence; RfD: Tariff Methodologies: 2 nd Edition on Storage & tariffs and the 5 th edition on the setting of Pipeline tariffs: 31 March 2011. Page 2 of 5

- incorporating stakeholder suggestions to align the cost of capital calculations more closely with local corporate finance practices; and - simplifying the tariff methodologies. Process of Public Participation 6. On 24 February 2011 the proposed changes to the Methodologies were posted on NERSA s website for inviting the public to comment on them at a public hearing scheduled for 15 th March 2011. A notice to this effect was published in the following newspapers on 24 February 2011: - The Business Day - The Star - Cape Times - Pretoria News - Natal Mercury 7. The public hearing was duly convened on 15 March 2011. Only Transnet Limited (Transnet) requested an opportunity to make representations. Transnet was, in general, pleased with the proposed refinements and made suggestions for further clarifications and refinements in certain areas. Stakeholder comments 8. The only comments received by NERSA on the proposed changes to the tariff methodologies were those presented by Transnet at the public hearing. 9. Comments presented by Transnet: RfD: Tariff Methodologies: 2 nd Edition on Storage & tariffs and the 5 th edition on the setting of Pipeline tariffs: 31 March 2011. Page 3 of 5

i. Transnet is in support of the use of the following in the tariff methodology: - an unadjusted risk free rate in determining the return on equity (Ke); - the ALSI in determining market returns; - market values for delivering equity betas. ii. Transnet proposed the following refinements to the tariff methodologies: - measure beta of proxy companies on a weekly basis over a period of 2 years; - a correction to the beta formula - no need to change to the Harris and Pringle formula when adjusting the beta of proxy companies; 10. NERSA s experience has been that the supply of data on such matters as company betas can change over time and that the Methodologies should not be burdened with amendment each time data providers change or make a change. Good practice also requires that the frequency of data measurements used for beta calculations should match the frequency of data measurements used in the calculation of the market risk premium. There are thus two data sources to be considered. Accordingly NERSA will clarify such matters in its Frequently Asked Questions. 11. NERSA does not agree with Transnet that the beta formula proposed by NERSA was incorrect. 12. NERSA is persuaded that its arguments in favour of a change to the Harris and Pringle formula outweigh the advantages Transnets sees in maintaining the Hamada formula. RfD: Tariff Methodologies: 2 nd Edition on Storage & tariffs and the 5 th edition on the setting of Pipeline tariffs: 31 March 2011. Page 4 of 5

Documents 13. The changes to the tariff methodologies and the reasons for these changes are presented in table format and is attached as Annexure A. Conclusion 14. The refinements made to the Methodologies in this decisions are warranted and in compliance with the Act and the Regulations. ************************ RfD: Tariff Methodologies: 2 nd Edition on Storage & tariffs and the 5 th edition on the setting of Pipeline tariffs: 31 March 2011. Page 5 of 5

Table 1: Summary of in the Tariff Methodologies: 5 th Edition: Tariff methodology for the setting of Petroleum Pipelines 2 nd Edition: Tariff methodology for the approval of Petroleum Facilities and Petroleum Storage Facilities. Pipelines Methodolog y Storage & Methodolog y Item Refinement 5.5 Minimum debt level 5.5 Minimum debt level Assumed minimum debt level It will be assumed that the licensee will have a minimum gearing level of 30%. It will be assumed that the licensee will have a minimum debt to total capital level of 30%. 5.6.3 WACC (Ke) 5.6.3 WACC (Ke) Cost of Equity (Ke) Formula Ke formula: 300 months Rf m m 1 Ke real 300months 300 months m m 1 300 months Rm - Rf CRA * Ke = (Rf +CRA)+(MRP * beta) + SSP+α+LP 5.6.3 - WACC (Rf) 5.6.3 - WACC (Rf) Risk-free rate Rf post-tax, real Rf after-tax real (storage wording). Rf - real. 5.6.3 - WACC (CRA) 5.6.3 - WACC (CRA) Country Risk Adjustment (CRA) CRA added to MRP CRA added to Rf, not MRP. 5.6.3 - WACC (MRP) Storage MRP market proxy JSE Financial and Industrial Index JSE All Share Index (ALSI). Page 6 of 25

Methodolog y Storage & Methodolog y uses All Share Index (ALSI) Item Refinement (FNDI) 5.6.3 - WACC (Beta) Does not pertain to storage Number of Proxy companies 6 Pipelines companies At least 6 Pipelines companies. Note 3 Beta Does not apply Formula for de-levering and re-levering Hamada formula Harris and Pringle formula. Note 3 - Beta Does not apply Market value versus book value Not specified but NERSA uses book values in practice. Market values when available, otherwise book values. 5.7 Cost of debt 5.7 Cost of debt Cost of Debt (Kd) Conversion of Kd from nominal to real: CPI forecast as at twelve months before the commencement of the tariff period under review is used to convert cost of debt from nominal to real terms. Conversion of Kd from nominal to real: The most recent CPI forecast, as at time of tariff setting, will be used to convert cost of debt from nominal to real terms. 7 - Tax Does not apply Tax example Only one example provided for both normalised and flow-through taxes Example will be removed from methodology. Two examples will be provided in FAQs one for flow-through tax and another for normalised tax. Page 7 of 25

Methodolog y Storage & Methodolog y Item Refinement 10.3 Operating efficiency Does not apply Operating Efficiency Adjustment Silent on the treatment of the difference between estimated operating expenditure and actual operating expenditure First claw back or give back the difference between estimated operating expenditure and actual operating expenditure. Then apply Operating Efficiency Adjustment as in Tariff Abbreviations and Acronyms Abbreviations and Acronyms List of Abbreviations and Acronyms List of abbreviations and acronyms Changes came about because of standardisation of abbreviations and acronyms used in the text of the tariff methodology. Page 8 of 25

Table 2: Detailed notes on the to the tariff methodologies 2 nd and 5 th Editions. 5 th Edition: Tariff methodology for the setting of Petroleum Pipelines 2 nd Edition: Tariff methodology for the approval of Petroleum Facilities and Petroleum Storage Facilities. Pipelines 1. Weighted Average Cost of Capital (WACC) 5.5 Minimum debt level 5.5 Minimum debt level It will be assumed that the licensee will have a minimum gearing level of 30%. It will be assumed that the licensee will have a minimum debt to total capital level of 30%. Clarity is given on the minimum debt level. 5.6.3 Cost of Equity (Formula) 5.6.3 Cost of Equity (Formula) Ke formula: Ke real 300 months m 1 300months 300 months m m 1 300 months Rf m Rm - Rf CRA * Where K e real = Post tax real cost of equity Ke formula: Ke = (Rf +CRA)+(MRP * beta) + SSP+α+LP Where: Ke = Post-tax real cost of equity {the other elements are explained below} The formula change is intended to align the methodology more closely with the conceptual approaches in corporate finance and economics disciplines as currently applied by economists and corporate finance practitioners in South Africa; The new formula also helps in simplifying the methodology. 5.6.3 Cost of Equity (Rf) 5.6.3 Cost of Equity (Rf) Rf = Post-tax risk free rate Rf = The pre-tax real risk free rate of interest. This is the average of the real monthly marked-to-market market practice supports the use of a pre-tax Rf; Rf is a measure that indicates what the Page 9 of 25

riskfree rate for the preceding 300 months for all government bonds 1 with at least a 10 year maturity as at twelve months before the commencement of the tariff period under review. sovereign risk is and hence should not be adjusted for tax purposes; Convergence with NERSA methodologies in other industries. Note: Formula of converting Rf from nominal to real terms will be removed. All calculations will be provided in FAQs such as the one already posted on the NERSA website 5.6.3 Cost of Equity (CRA) 5.6.3 Cost of Equity (CRA) CRA = country risk adjustment CRA is added to MRP CRA = Country Risk Adjustment The pre-tax, real CRA will be added to the risk free rate. The CRA is for assets in another country outside South Africa that are an integral part of the same assets within South Africa. Country risk is reflected in the sovereign risk free rate, i.e. the country risk is normally reflected in the government debt which is reflected in the proxy Rf; Hence the appropriate place to adjust for country risk is in the Rf rather than in the MRP; Country Risk Adjustment relates to non-south African country risk as South African country risk is already reflected in the risk free rate. The main components of country risk are 1 Data on government bonds are sourced from the South African Reserve Bank and published by NERSA. Page 10 of 25

The adjustment is for that other country concerned. political risks, infrastructure risks, corruption risks and inherent country risk associated with the country s economy; Given that NERSA regulates South African assets, it is probable that few, if any, of the non-south African risk factors will be present in an asset being regulated by NERSA. It is recommended to adjust the risk free rate where an adjustment for country risk is required. 5.6.3 Cost of Equity (MRP) Storage uses All Share Index (ALSI) MRP = the post tax, real market risk premium. The proxy used is the JSE Financial and Industrial Index (FNDI) MRP = the post-tax, real market risk premium. The proxy used for the market return is the Johannesburg Stock Exchange (JSE) All Share Total Return Index (ALSI) for the preceding 300 months as at twelve months before the commencement of the tariff period under review. Data for the FNDI is only available for 15 of the 25 years and the Energy Regulator therefore has to artificially construct a FNDI index (which creates practical and theoretical difficulties); To achieve consistency with Electricity 2 and Piped-gas 3 industries regulated by NERSA which are already using ALSI in determining the market risk premium value Note: Formulas for determining 2 the Decision on Eskom Holdings Limited: Revenue Application Multi Year Price Determination 2010/11 to 2012/13 (MYPD2), par 103-110. 3 ALSI was applied in the Energy Regulator decision on Transnet Pipelines gas transmission tariff for the Lilly pipeline for 2010, 31 August 2010. Page 11 of 25

the MRP as well as of converting Market Returns (MR) from nominal to real terms are removed. All calculations will be provided in FAQs such as the one already posted on the NERSA website. 5.6.3 Cost of Equity (Beta) Does not pertain to storage = Beta, the systematic risk parameter for regulated entities providing pipeline, storage and loading facility services. The beta must be determined by proxy. As a proxy the average of six (6) pipeline companies listed on stock exchanges must be used. The methodology to be used to determine the beta is set out in Note 3. = Beta, the systematic risk parameter for regulated entities providing pipeline, storage and loading facility services. The beta must be determined by proxy. As a proxy the average of at least six (6) pipeline companies listed on stock exchanges must be used. The methodology to be used to determine the beta is set out in Note 3. To provide for the fact that at times there are industry consolidations/ takeovers which means that beta for a proxy company that is taken over ceases to exist. Providing for a larger list enables NERSA to still have the requisite minimum of 6 proxy companies from which to calculate the asset beta; Companies operating in the petroleum pipeline industry are often involved in more than just petroleum pipeline operations. Often their operations include storage and loading activities; In selecting proxy companies for the beta analysis there is invariably a trade-off between achieving a large, statistically reliable sample, and restricting the sample to comparable companies whose characteristics closely match those of the company for which the cost of capital is being calculated. Note Page 12 of 25

NERSA will continue to publish the names of those companies once per annum. 5.6.3 New Cost of Equity Small stock premium (SSP) 5.6.3 New Cost of Equity Small stock premium (SSP) Included in Beta as part of the beta adjustment. SSP = Small Stock Premium. Size of the licensed entity if the circumstances warrant such an adjustment. This tends to follow a qualitative analysis process, but will be informed by authoritative publications on prevailing practices by local corporate finance practitioners. Previously provided for in beta adjustment, but now shown separately so as to leave beta as a pure measure of systematic risk; On average, smaller companies have higher rates of return than larger companies. Empirical evidence in support of size adjustments to the cost of equity also exists; However, in the context of setting pipeline tariffs, many of the risks that smaller companies may face, such as lack of access to finance or lack of management expertise may not be present in the company being regulated i.e. this adjustment will only be applied if the circumstances warrant it. 5.6.3 New Cost of Equity Project specific risk(α) 5.6.3 New Cost of Equity Project specific risk(α) Included in Beta as part of the beta adjustment. α =Project specific risk if the circumstances warrant such an adjustment. Previously provided for in beta adjustment, but now shown separately so as to leave beta as a pure measure of systematic risk; In the presence of uncertainty and sunk costs, investment thresholds that exceed the WACC are required to enable investment. The WACC just reflects the systematic risk. It does not reflect the probability of bankruptcy or other Page 13 of 25

idiosyncratic risks that firms prudently consider when making investment decisions 4. Empirical evidence showing that specific risks are considered in project appraisals exists. However, these risks may not be relevant in a regulated environment where the project s returns are certain i.e. this adjustment will only be effected if the circumstances warrant it. 5.6.3 New Liquidity Premium (LP) 5.6.3 New Liquidity Premium (LP) Included in Beta as part of the beta adjustment LP = liquidity premium to accommodate companies which are not publicly traded if the circumstances warrant such an adjustment. Previously provided for in beta adjustment, but now shown separately so as to leave beta as a pure measure of systematic risk; It is generally accepted in financial theory that equity investments that are not publicly traded sell at a discount relative to equity investments that are publicly traded. 2. Beta (Note 3 for the Determination of Beta) NOTE 3 Step 1 Does not apply Beta calculation Hamada formula which includes the tax shield in the notation is used. The following formula must be used to determine the asset beta: Beta calculation The Harris and Pringle formula which excludes the tax shield in the notation will be used. The following formula must be used to current specifies use of the Hamada formula to adjust the betas for financial leverage. The Hamada formula assumes the future debt servicing schedules are known with certainty. However, treating companies future debt tax shield as certain is not always a realistic assumption to make; 4 Lewis Evans and Graeme Guthrie, Efficient Price Regulation of Networks that have Sunk Costs, August 2002 Page 14 of 25

a1 Where: 1 1 Dt 1 Tr* Eq β a 1 = Asset beta for company proxy 1 β1 = Beta of company proxy 1 (will be given) = Tax rate of relevant country D t = Debt = ERquity T r E q Repeat step 1 for each of the 6 chosen proxy firms. determine the asset beta β a1 = β1 / [1 + Dt/Eq] Where: β a 1 = Asset beta for proxy company 1. β1 = Beta of proxy company 1 Dt = Debt of proxy company 1. Eq = Equity of proxy company 1. Repeat step 1 for each of the 6 or more chosen proxy firms. Instead we recommend using the Harris and Pringle formula for beta un-levering and relevering as this formula considers interest tax shields to be uncertain. Companies are likely to change gearing levels in response to changes in enterprise value; The Harris and Pringle formula also enhances the simplicity of the calculation; The effect of substituting the Harris and Pringle formula for the Hamada formula will be pronounced only if the proxy companies and the subject company have materially different tax rates. Note: The steps for de-levering and re-levering remain as before in the. Only the formula changes from the Hamada formula to the Harris Pringle formula. NOTE on un-levering and re-levering Does not apply Beta calculation:un-levering Book values are used in delevering. Beta calculation: unlevering. Market values for proxy Since beta is calculated using market values, some stakeholders have expressed preference for using book values notwithstanding the limitation that not all entities have market Page 15 of 25

companies will be used where such market values exist. Where no market values exist for proxy companies, book values will be used values and market value of debt is not readily available; Moreover, other regulators have concluded that there is no difference in results if either of the two is used in an internally consistent manner 5. NERSA concurs with this conclusion and will use market values where such values exist. Where there are no market values, then book values will be used. To deal with this uncertain aspect, NERSA will address the use of market value versus book value when determining debt /equity ratios for de-levering and re-levering in the FAQs. This way the calculation will be done to reflect the market conditions and availability of market or book values at the time of calculation for the respective entity for whom the tariff is being set. NOTE 3 Step 2 Does not apply No change N/A Note 3 Step 3 Does not apply Calculation of beta (β) for licensee The following formula must be Calculation of beta (β) for licensee The following formula See above reasons for switching from the Hamada formula to the Harris and Pringle formula. 5 Australian Energy Regulator, Final decision: Electricity transmission and distribution network services providers, Review of the weighted average cost of capital (WACC) parameters, May 2009. Page 16 of 25

used to determine the beta for the licensee: BL = WA β * (1 +(1-Tr)* ((Dt/Eq))) must be used to determine the beta for the licensee: B L = β a E * [1 + D t /E q ] Where: Β L = Beta for the licensee β a E = Weighted average β of the proxy firms asset betas from Step 2. D t E q = Interest bearing debt of the licensee subject to a minimum gearing level of 30%. = Equity of the licensee. Note 3 Beta adjustments Does not apply Adjustments to beta possible for company size and company specific risk. No beta adjustments. Instead adjustments may be made directly to the cost of equity (Ke) for SSP, LP and alpha as explained above. Best practice is not to adjust the beta but to apply overall adjustments to the cost of equity as set out in the proposed Ke formula above; This is therefore to align with corporate finance practitioners in South Africa who do not adjust beta; Beta is not adjusted by these practitioners in Page 17 of 25

order to maintain its purity as a measure of systematic risk that cannot be diversified away by an investor; Adjustments in return to reflect licenseespecific risk factors should therefore be effected by separate changes to the Ke. Such license-specific risk factor adjustments include; size, liquidity and project specific risks. 3. Consumer Price Index forecast. 5.7. Cost of Debt. 5.7. Cost of Debt. Consumer Price Index forecast (CPI f ) used in the cost of debt formula CPI f as at twelve months prior to the commencement of the tariff period under review is used to convert nominal cost of debt to real cost of debt The latest available Consumer Price Index forecast (CPI f ) as at the time of setting the tariffs will be used to convert nominal cost of debt to real cost of debt Forecasts made further back from the actual date tend to be less accurate than those made closer to actuals date, which has the consequence of larger claw back adjustments. Large claw-back adjustments increase tariff uncertainty. Therefore, using the most recent available CPI forecasts minimises the claw-backs (since they are nearer to the actual date) and thereby engenders tariff certainty. Note NERSA will continue to source CPI forecasts from the Bureau for Economic Research (BER). 4. Tax calculation example. 7 Tax (T) 7 Tax (T) Tax Tax Page 18 of 25

In the current tariff methodology there is only one example of how to calculate tax. This example shows how tax is calculated on a notional tax basis. Note: The example is provided in separate table (Table 4a) at the end of this table. The tax calculation example will be removed from the methodology and included in the FAQs and replaced by two examples. The two examples to be provided in the FAQs (see Note below) show how to calculate tax using the following two approaches: i. Flow- through tax ; and ii. notional tax Note: The examples are provided in separate tables (Table 4b and 4c) at the end this table. Removing examples from the methodology allows the methodology to focus on the principles of tariff setting Detailed mechanics are clarified elsewhere; This separation also lightens the administrative burden of making improvements to examples. If examples were included in the methodology, any small improvement would entail a burdensome process of having to amend the ; In the current methodology, the example on the calculation of tax is based on formulas that do not deal with all elements for both flow-through and notional tax approaches. The two examples in the proposed amendment are based on the expanded formulas that deal with all elements and specifically with the tax shield on the amortisation of the write-up. 5. Operating efficiency adjustment. 10.5 Minimum Debt Level Does not apply Silent on the treatment of the difference between estimated operating expenditure and actual operating expenditure First claw back or give back the difference between projected operating expenses and Clarify the calculation of Operating Efficiency Adjustment by showing the first step in the calculation i.e. the difference between estimated and actual operating expenses. Page 19 of 25

actual operating expenses.. Then apply Operating Efficiency Adjustment as in Tariff. Apply the following formula to calculate clawback on operating expenses. OpexA= Opexp-Opexa Where: OpexA = Operating Expense Adjustment Opexp = Operating and maintenance expense projected Opexa = Operating and maintenance actual. Then apply the following operating efficiency adjustment formula: OeA = (EaOC AaOC) x actual volumes x 50% Where: OeA = Operating efficiency adjustment. AaOC = Actual average operating costs (Actual operating expenditure Page 20 of 25

volumes). EaOC = Estimated average operating costs (estimated operating expenditure estimated volumes). Abbreviations and Acronyms And editorial corrections Abbreviations and Acronyms and editorial corrections Editorial changes and changes to the abbreviations and acronyms used in the 5 th Edition of the Tariff Methodologies as well as editorial corrections. Throughout the document some editorial changes were made to improve on the general layout of the document. Changes to the abbreviations and acronyms were effected to have a standardised format and structure used for abbreviations and acronyms. Page 21 of 25

Table 4a: Tax table. : The table below provides the example of a tax calculation for notional tax. Table 4a:Tac in Example Reference Year 1 Ke a 5.10 Kd(real) b 1.78 WACC c=a+b 6.88 E d 3.00 D e 4.20 F f 0.00 C g 0.00 Allowable revenue before tax allowance h=c+d+e+ f+g 14.08 Actual Interest cost (nominal) i 5.88 Tr j 28% NPBT excl tax allowance={(rab*wacc)+e+d (historic)+f±c}-{e+d(historic)} [Note 1: interest not deducted to allow Tax shield] [Note2: The licensee must, before gross-up, adjust only the k=h-d-e subtraction leg -{E+D} to use the tax deductible amounts allowed by the tax authority] 6.88 NPBT excluding tax allowance k Tax *Tr 1- Tr L * j 1- j 2.68 Total Allowable revenue m=h+l 16.76 Taxable income notional formula n=k+l 9.56 Taxation paid notional formula o=n*j=l (2.68) NPAT p=n-o 6.88 Actual tax Calculation Taxable income q=m-d-e-i 3.68 Taxation paid Actual r=q*j (1.03) Page 22 of 25

Tax Shield (benefit to investor) s=o-r 1.65 Actual interest t=i 5.88 Tax shield proof u=t*j=s 1.65 All the elements in the tax formula are not covered in the above table, and is therefore replaced. It will be replaced by two separate tables, one applicable for notional tax and another for flow-through tax. Note: These tables will not appear in the Tariff, but will be published in the FAQ-document which is also published on NERSA s website. Page 23 of 25

Table 4b: Example on how to calculate the Flow through tax Formula to apply when a flow-through tax approach is followed: Tax = {(NRBTA) / (1-t)} *t Where: NRBTA = Net revenue before tax allowance = {(RAB*WACC)+E+D(historic & write up)+f± C} - {E+wear and tear (historic) +Kd(nominal)}. Example: Flow through taxes Ke*RAB*Equity ratio a 5.10 Kd*RAB*Debt ratio b 1.78 WACC*RAB c=a+b 6.88 E d 3.00 D (historic) e 4.00 D (write-up) ee 0.20 Wear and tear allowance eee 10.00 F f - C [Claw back Excluding Tax claw back] g - NRBTA=Allowable revenue before tax allowance h=c+d+e+ee+f+g 14.08 t j 28% Kd (nominal) n 5.88 NRBTA={(RAB*WACC)+E+D(historic & write up)+f+-c}-{e+wear & tear (historic) +Kd(nominal)} Note that interest is deducted as a actual tax calculation is performed k=h-d-eee-n (4.80) Tax={(NRBTA)/(1-t)}*t l={k/(1-j)}*j (1.87) Total Allowable revenue m=h+l 12.22 Test tax rate l/(k+l) 28% Note that any other non-taxable income or non-deductible expenses have not been included above. These will form part of the actual calculation when detail is available. The example on notional tax is provided on the next page. RfD: Tariff Methodologies: 2 nd Edition on Storage & tariffs and the 5 th edition on the setting of Pipeline tariffs: 31 March 2011 Page 24 of 25

Table 4c: Example on how to calculate Notional tax Formula to apply when a flow-through tax approach is followed: Tax = {(NRBTA) / (1-t)} *t Where: NRBTA = Net revenue before tax allowance = {(RAB*WACC)+E+D(historic & write up)+f± C} - {E+D(historic)}. Example: Notional Tax Ke*RAB*Equity ratio a 5.10 Kd*RAB*Debt ratio b 1.78 WACC*RAB c=a+b 6.88 E d 3.00 D (historic) e 4.00 D (write-up) ee 0.20 Wear and tear allowance eee 10.00 F f - C [Claw back Excluding Tax claw back] g - NRBTA=Net revenue before tax allowance h=c+d+e+e e+f+g 14.08 t j 28% Kd (nominal) n 5.88 NRBTA={(RAB*WACC)+E+D(historic & write up)+f+- C}-{E+D (historic) } Note that interest is not deducted to allow for the tax shield on after tax real Kd. k=h-d-e 7.08 Tax={(NRBTA)/(1-t)}*t l={k/(1- j)}*j 2.75 Total Allowable revenue m=h+l 16.83 Test tax rate l/(k+l) 28% RfD: Tariff Methodologies: 2 nd Edition on Storage & tariffs and the 5 th edition on the setting of Pipeline tariffs: 31 March 2011 Page 25 of 25