The assumptions used in valuing the pension liability are both judgmental and sensitive to change and thus there is a risk that a small change in the judgements used will have a significant impact on the valuation of the pension deficit.

The continued fall in gilt rates and the low yield environment has reduced the discount rate on which the Group’s pension deficit is calculated to 2.7% (2015: 3.9%), while the remaining assumptions are reasonably consistent with the prior year.

As such our area of focus was on the assumptions used in calculating the liability, particularly the discount rate.

Refer to pages 76 and 77 for details on the Audit Committee reviews and conclusions and notes 3 and 22 in the Financial Statements.

We used PwC pension specialists to help us assess the assumptions used by the Directors in the valuation of the pension deficit. We compared the discount and inflation rates used in the valuation to our internally developed benchmarks. We have an internally developed range of acceptable discount rates for valuing pension liabilities, which is based on our view of various economic indicators. While our range is, itself, subjective, the discount rate used by the Group is in the middle of our expected range, however the inflation RPI is at the more optimistic end of the range.

Based on the work performed, we did not identify any material issues over the assumptions used in valuing the pension deficit.

The Group has £5.3 billion of property, plant, and equipment, the majority of which relates to gas production, storage and power generation assets, £1.8 billion of intangible assets and £2.6 billion of goodwill.

Impairment assessments of these assets require significant judgement and there is the risk that valuation of the assets may be incorrect and any potential impairment charge or reversal miscalculated.

The value of Centrica’s assets is supported by either value in use calculations, which are based on future cash flow forecasts or fair value less costs of disposal.

There has been some improvement in forecast oil and gas prices in 2016. In addition, favorable oil and gas reserve revisions and cost reductions have resulted in pre-tax impairment reversals of £63 million in relation to the Norwegian gas and oil assets in the Exploration & Production business. A pre-tax impairment reversal of £56 million has been recognized in relation to certain gas assets in Trinidad and Tobago after considering the proceeds expected under the sale and purchase agreement. An amount of £16 million (pre-tax) was also reversed in relation to decommissioning on previously impaired assets.

An impairment assessment was performed on combined cycle gas turbine (CCGT) power stations with a pre-tax impairment reversal of £26 million recognized within exceptional items in relation to the King’s Lynn power station following the award of a 15 year capacity market contract.

Impairment indicators were identified for the Storage facility following operational issues. This has resulted in a total pre-tax impairment charge of £176 million being recognized. The model remains highly sensitive to key assumptions including price, volume and well performance.

In 2015 the Group announced E&P Canada was no longer a core part of the Exploration & Production business and would be divested.

The Directors have received a range of bids and have considered internal discounted cash flow analysis which evidences a wide range of possible outcomes in relation to the valuation of the business.

Considerable judgement has therefore been applied in determining the recoverable amount of the assets. As significant uncertainty remains in the disposal process for E&P Canada, the Directors were unable to conclude that completion of the sale is highly probable within 12 months from balance sheet date. Therefore it has not been classified as held for sale at year end.

Refer to pages 76 and 77 for details on the Audit Committee reviews and conclusions and notes 3, 7, 13, 15 and S2 in the Financial Statements.

We assessed and challenged the impairment analysis prepared by the Directors as outlined below:

With regard to the overall impairment assessments performed by the Directors, we evaluated the design of internal controls in place to check that the Group’s assets are valued appropriately including those controls in place to determine any asset impairments or impairment reversals. We also reviewed the assets that Directors assessed for indicators of impairment and no indicators were identified.

We evaluated the Directors’ assumptions and estimates used to determine the recoverable value of the gas and oil production and storage assets, power generation assets, intangible assets, and goodwill. This included reviewing fundamental curves, benchmarking their oil and gas price assumptions, reviewing operating cost forecasts and expected production profiles. We tested these assumptions by reference to third party documentation where available, such as commodity price forecasts, and consultation with operational management. With regard to Trinidad and Tobago, we reviewed the Directors’ assessment of the recoverable amount of the assets with consideration to the sale and purchase agreement.

We used PwC valuation specialists to help us assess the commodity prices and discount rates used by the Directors. We benchmarked these to external data and challenged the assumptions based on our knowledge of the Group and its industry. In addition we tested the Directors’ sensitivity and stress test scenarios to ensure appropriate judgement had been applied.

We challenged the key assumptions used in each impairment model and performed sensitivity analysis around key drivers of cash flow forecasts, including output volumes, commodity prices, operating costs and expected life of assets.

With regard to E&P Canada, we have challenged the Directors’ assessment that the business should not be classified as held for sale at the balance sheet date. This included meeting with operational and group management to assess the status of the disposal process as at the balance sheet date, understanding the status of bids, negotiations and assessment of the options being considered by the Directors. In relation to assessing the recoverable amount of the E&P Canadian assets, we have considered the Directors’ internal discounted cash flow analysis, performed a range of sensitivity analyses on the discounted cash flow and understood the third party bids received.

Based on our analysis and the analysis performed by our valuations team, we did not identify any material issues with the impairment conclusions and valuation of exploration and production, storage and power generation assets and goodwill. We did not identify any material issues with the accuracy of the impairment charges and reversals and the associated disclosures including the classification of assets

The Group enters into a number of significant and complex contracts, for example, power procurement and tolling contracts. Macro-economic factors, such as forecast commodity prices, can have a significant impact on the profitability of these contracts, and therefore the Directors make an assessment as to whether the impact of such factors has resulted in contracts becoming onerous.

The onerous contract provision of £64 million in relation to the Spalding power station tolling agreement has been released following the renegotiation of the contract and further improvements in elements of the cash flow forecasts underpinning the provision. The Group has also negotiated an exit from the Rijnmond tolling contract which resulted in an additional onerous contract provision of £26 million being recognized.

The onerous contract provision for the North America wind farm power purchase agreement was updated to take into account changes to forecast US power prices resulting in a £15 million release to the provision being recorded at year end.

The Directors’ existing assessment of expected costs in relation to the European gas transportation contract remains materially unchanged.

Our focus on onerous contracts was assessing whether material onerous contracts have been identified and that the valuation of any provision is materially correct.

Refer to pages 76 and 77 for details on the Audit Committee reviews and conclusions and notes 3, 7 and 21 in the Financial Statements.

We tested the identification and completeness of onerous contracts through discussions with management, examination of board minutes, obtaining and reading the new significant contracts entered into during the year and testing Directors’ assumptions for a sample of contracts.

We tested the valuation of the onerous contract provisions by evaluating whether appropriate judgements and assumptions had been applied in determining the unavoidable costs of meeting the obligation and the estimate of the expected benefits to be received under the contract.

For the Spalding onerous contract provision release, we examined the revised contract and tested the updated cash flow forecasts. In relation to the Rijnmond tolling contract settlement, we have reviewed the settlement agreement and verified the cash payment made. We have also reviewed management’s updated model for the North America wind farm power purchase agreement to ensure the appropriate judgement had been applied.

The Group enters into a number of forward energy trades to help protect and optimize the value of its underlying production and storage assets, power generation assets, and transportation assets, as well as to meet the future energy and supply needs of customers.

Certain of these arrangements are accounted for as derivative financial instruments and are recorded at fair value.

Judgement is required in valuing these derivative contracts, particularly where the life of the contract is beyond the liquid market period. The fair value calculation requires bespoke models to be used that are specific to the derivative and, as such, we gave particular focus to the valuation of derivative contracts at the balance sheet date.

Refer to pages 76 and 77 for details on the Audit Committee reviews and conclusions and notes 2 and 7 in the Financial Statements.

We assessed the overall commodity trading process including internal risk management procedures and the system and controls on the origination and maintenance of complete and accurate information relating to derivative contracts. We found the controls in place over this process to be operating effectively and therefore placed reliance on these controls in our testing.

We tested the valuation of derivative contracts at the year-end date which requires the use of the Directors’ valuation models. Our audit procedures focused on the integrity of these valuation models and the incorporation of the contract terms and the key assumptions, including future prices and discount rates. We verified input prices into the system and recalculated valuations for a sample of derivatives, as well as performing sensitivity analyses for more complex derivatives. Our testing focused on ensuring appropriate judgement had been applied in the valuation of the contracts and we did not identify any material errors.

The middle column of the Income Statement represents exceptional items and certain re-measurements. In the current year, there is a total pre-tax exceptional charge of £11 million and a £1,030 million pre-tax net gain relating to re-measurements, included within operating profit.

Exceptional items

The current year exceptional items pre-tax charge comprises of restructuring costs of £228 million and an impairment charge on the UK gas storage assets of £176 million. Impairment reversals were recognised on certain exploration and production assets of £135 million and combined cycle gas turbine (CCGTs) power stations of £26 million.

A net gain on disposal of businesses and assets of £101 million, a one-off past service credit as a result of the implementation of a reduced salary cap on pensionable pay of £78 million, and a net release of onerous power procurement contracts of £53 million were also included within exceptional items in the current year.

The appropriate classification of exceptional items involves subjective judgement by the Directors including whether the item is truly exceptional by virtue of its nature, size or incidence. Our focus was on testing that the presentation and disclosure of these items is materially correct.

Certain re-measurements (as defined in the Financial Statements) Certain re-measurements which resulted in a pre-tax net gain of £1,030 million, relates to the fair valuing of forward energy trades. There are two main types of trades the Group participates in:

• optimisation trades – it is the Directors’ view that movements in the fair value of optimisation trades do not reflect the underlying performance of the business because they are economically related to parts of the business which are not fair valued, for example exploration and production assets or downstream demand. As such, these trades are only reflected in business performance when the underlying transaction or asset impacts the profit or loss; and

• speculative trading – it is entered into for the purpose of making profit. Therefore all fair value movements associated with it are disclosed as part of underlying business performance. Our focus was on testing the correct classification of optimization and speculative trades.

Refer to pages 76 and 77 for details on the Audit Committee reviews and conclusions and notes 2 and 7 in the Financial Statements.

For each of the material exceptional items, we considered Directors’ analyses of why they were determined to be exceptional and performed our own, independent assessment by looking, primarily, at the nature of the items. The detailed work we performed on the exceptional items relating to the impairment charges and reversals, which is one of the most significant items, is described on pages 126, 127 and 128.

We have performed audit testing over the restructuring costs recorded during the year including assessing whether any related provisions meet the requirements under IAS 37 Provisions, Contingent Liabilities and Contingent Assets. Our testing did not identify any material issues and we have ensured appropriate judgement has been applied in classifying restructuring costs as an exceptional item.

For certain re-measurements we audited the principles management use to determine whether a trade should be recognised as part of business performance or presented separately. We evaluated whether the agreed principles had been applied consistently by testing that a sample of the trades had been presented correctly as optimisation or speculative trading. Based on the work performed we did not identify any material issues with the presentation, classification or disclosure of exceptional items and certain re-measurements.

The accuracy of the recorded energy services revenue within the Group and its presentation in the income statement is dependent on complex estimation methodologies and algorithms used to assess the amount of energy supplied to customers between the date of the last meter reading and the year end (known as unread revenue). Unread gas and electricity revenue comprises both billed and unbilled revenue. The specific risk over unread revenue is the accuracy of the estimation. Where an unread estimate is billed this gives the customer an opportunity to challenge the estimate which can lead to the subsequent refinement of unread estimates. Where unread estimates are unbilled, there continues to be a risk over accuracy, recoverability and therefore correct recognition in the income statement and balance sheet.

Furthermore, following the implementation of a new billing system in UK Business in 2014, Directors’ have performed additional levels of review over the revenue and receivables cycle including making judgements over the level of accounts receivable provisioning.

Refer to pages 76 and 77 for details on the Audit Committee reviews and conclusions and notes 3, 4 and 17 in the Financial Statements.

In order to test the accuracy of the unread billed and unbilled revenue in UK Home, UK Business and UK Home Services, we assessed the IT general controls, system application configuration, and business process controls in relation to the revenue estimation and billing systems.

Our testing in these areas was sufficient to enable us to place reliance on the system generated revenue estimation for the year end audit. In North America Home and North America Business, we performed detailed testing to support the accuracy of the unread billed and unbilled revenue.

Given the relatively short time period between the end of the financial year and the audit, the majority of unbilled revenue as at 31 December remained unbilled and uncollected at the date of this report. We therefore focused our substantive testing on the manual adjustments to estimated unbilled revenue, assessing the appropriateness of the estimation methodologies and the reconciliation of unbilled reports to the general ledger at the year end. Where manual adjustments were made to the unbilled revenue estimate, we challenged the basis of the adjustments made, the source of the data used and the consistency of the adjustments with prior years to confirm we were comfortable with the adjustments.

In assessing the methodology used to derive the unbilled revenue estimate at the balance sheet date, and testing the performance of historical billing and collections, we did not identify any material issues with the recognition of unbilled revenue.

With regard to the new billing system in UK Business, we increased our scope of work in order to assess any continued impact of the implementation, specifically on accounts receivable and associated provisioning. This included assessing the recoverability of debt and additional procedures over the calculation of the debt provision at year end.

Based on our work we did not identify any material misstatements with downstream revenue recognition.

National Grid provides defined pension and other post-employment benefits to employees in the UK and US through a number of schemes. At 31 March 2016, National Grid’s gross defined benefit obligation is £29.0bn which is offset by scheme assets of £26.4bn which are significant in the context of both the overall balance sheet and the results of the Group.

The valuation of the pension liability requires significant levels of judgement and technical expertise in choosing appropriate assumptions. Changes to the key assumptions including salary increases, inflation, discount rates, and mortality can have a material impact on the calculation of the liability.

Also, the pension plan assets include a number of investments for which there is no observable input to the fair value (i.e. no quoted market price); the valuation technique used to measure the fair value of these assets involves a number of subjective judgements.

We have tested the significant judgements made by National Grid’s third party actuaries as set out below and assessed their independence and competence. We found no material issues that would impact our audit approach.

We agreed the discount and inflation rates used in the valuation of the pension liability to our internally developed benchmarks. We compared the assumptions around salary increases and mortality to national and industry averages. All of the assumptions used fell within our acceptable range.

We obtained details of the measurement of fair value for assets with unobservable inputs. Such assets were typically private equity or real estate fund investments for which we obtained audited financial statements in support of the measurement of net asset value. We found no material issues from this testing.

National Grid US are going through a finance function reorganization and a programme of process and control improvements. In this period of change and until processes and controls are finalised and the new finance structure is embedded, there is higher risk of error in the financial information reported by the US Regulated business.

Change in level of risk year on year: No change

We are seeing progress in some processes on control remediation and additional focus on the control environment. Some areas, in particular property, plant and equipment (PPE) are proving more complex and will take longer.

As a consequence of the higher risk of error in financial information reported by National Grid US, a significant portion of both US and Group senior audit team members’ time has been spent developing our audit response to the US control environment, which is summarised below, and discussing this with management and the Audit Committee.

We performed additional testing of key account reconciliations across a number of different general ledger accounts, ensuring that significant reconciling items were supported with sufficient and appropriate documentation. Management continue to operate their additional control of preparing an aggregation of unreconciled items across all accounts in order to assess the potential impact of adjusting for these items. We tested this aggregation to ensure it was complete and accurate by agreeing these items to the underlying account reconciliations and vice versa. The net impact on the income statement if all unsupported reconciling items were to be resolved was below our reporting level for the Audit Committee.

We tested the design and operating effectiveness of journal review controls and found nothing that would cause us to believe these controls were not working as intended. We also tested manual journal entries based on a risk assessment of value and nature, with no matters arising that required reporting to the Audit Committee.

Over time National Grid has acquired, owned and operated a number of businesses that have created an environmental impact that will require remediation. This is particularly significant in the US partly as a result of National Grid’s exposure to certain ‘Superfund’ sites. At 31 March 2016, the total liability in respect of environmental provisions is £1.2bn, of which £0.9bn relates to the US.

Environmental provisions require significant judgement in determining the form of remediation and the timing and value of projected cash flows associated with it, including the impact of regulation, accuracy of the site surveys, unexpected contaminants, transportation costs, the impact of alternative technologies and changes in the discount rate.

Change in level of risk year on year: No change

In the US and UK, National Grid uses external and internal experts to help determine the total expenditure required to remediate sites. As part of the audit we obtained and inspected these experts’ reports and assessed their independence and competence and we found no material issues that would impact our audit approach.

For all material sites and a sample of other sites, we corroborated information on the nature of each of these sites to National Grid’s underlying site usage records. In addition, to assess the reliability of the experts’ estimates, we compared previous estimates against actual spend for sites which have been remediated, without material issue.

In the US, due to the individually significant sites, we utilized our own environmental specialists to review management’s key assumptions underlying the calculations. Where possible we confirmed other inputs into the calculation by reference to publicly available information and noted no exceptions.

We inspected responses to our confirmation requests from National Grid’s legal advisors in order to identify any issues related to the valuation of the Group’s exposure to environmental remediation costs and noted no issues.

In order to assess the reasonableness of management’s discount rate assumptions we compared these to our internally developed benchmarks, including performing sensitivity analysis. We identified a potential adjustment related to one discount rate which was marginally outside our expected range and reported this to the Audit Committee. We considered this immaterial for adjustment in the Group financial statements.

In order to fund its activities, at 31 March 2016 National Grid had total borrowings of £28.3bn, of which £6.4bn is denominated in currencies other than Sterling or US Dollars and exposes the Group to foreign exchange and interest rate risk.

As a result, the Group has a significant treasury operation with sophisticated risk management activities and uses financial instruments including complex derivatives to manage the foreign exchange and interest rate risks, primarily interest rate swaps and cross-currency interest rate swaps.

The valuation of a number of the derivative contracts entered into by National Grid is a complex and judgmental area and includes key assumptions over estimates of future interest, exchange rates and determination of appropriate discount rates to apply to future cash flows.

Change in level of risk year on year: Decreased

Because our work in 2014/15 on the introduction of regression for hedge effectiveness testing identified no issues.

We tested the design and operating effectiveness of IT General Controls including user access, change management and segregation of duties within the treasury management system and we found no material issues that would impact our audit approach.

We tested the design and operating effectiveness of key controls that relate to recording and valuing derivative transactions in the treasury management system. We also tested the accuracy and completeness of the information held within the system by agreeing to third-party confirmations and found no differences when compared to the system data.

We tested the models and key assumptions used by management to value complex derivatives which were agreed as appropriate. Where management entered into new significant contracts in the year, we tested the contracts and assumptions used to assess whether the accounting treatment adopted is in accordance with IAS 39.

A key focus for National Grid is network investment with total capital expenditures across the Group of £3.9bn during 2015/16 (2014/15: £3.5bn).

Depending upon its nature, expenditure may be capitalised as PPE or expensed in the year the cost is incurred. In making this decision the directors have to consider whether the expenditure will generate future economic benefits which necessarily involves judgment, for example in determining whether activities or items are adding value or maintaining existing assets.

In relation to the US, there was a heightened risk that the controls over: the classification of costs between PPE and expenses; and the controls over the transfer of assets under construction to assets in service may not have been working effectively due to control weaknesses previously identified. In addition, there are complex adjustments that are required to translate local plant accounting records prepared under generally accepted accounting principles in the United States (US GAAP) to comply with IFRS.

In relation to the UK, our interim review work highlighted weaknesses in relation to some reconciliations and classifications within PPE.

Change in level of risk year on year: Increased

Because of significantly increased level of capital expenditures in the US and the control weaknesses identified in the US and UK.

We assessed whether the Group’s accounting policies in relation to the capitalisation of expenditures complied with IFRSs, and tested the implementation of those policies through a combination of controls testing, including IT General Controls over the PPE accounting systems, and substantive testing of the supporting documentation behind the costs and we found no material issues that would impact our audit approach.

In the US, we performed additional testing to ensure the completeness and accuracy of capitalisation. Our procedures included identification of projects where the proportions of costs capitalised were different to those we would expect based on the nature of the work performed, and procedures around the appropriateness of capitalisation of payroll costs, noting that amounts had been properly recorded.

As a result of issues identified by our testing, we extended our sample of certain types of open work orders. Our work identified a small factual adjustment which we reported to the Audit Committee together with a larger projected adjustment. Taken together, these were considered to be not material for adjustment in the financial statements.

With respect to the IFRS adjustments to US GAAP reporting, we tested the analysis to underlying accounting records, recalculations and supporting documentation, identifying no adjustments that required reporting to the Audit Committee.

In the UK, we focused our testing on the capital expenditures that had the most significant value, with a particular focus on Electricity Transmission which is the largest area of UK capital expenditures. As part of our testing, we inspected contracts and underlying invoices to ensure the classification between capital and operating expenditure was appropriate. We reviewed the ageing of the assets under construction balance for indicators of impairment and key judgments associated with the PPE balance. Our approach is supported by comfort obtained from our testing of the key controls within the PPE process, which included reconciliations and controls over classification. Based on year-end tests we performed, the weaknesses identified at our interim review had been rectified by management. We found no material issues arising from this work.

In November 2015, National Grid announced its intention to dispose of a majority share of the UK Gas Distribution (UKGD) business. This will be a significant transaction as UKGD comprises approximately 21% of Group profit/net assets and in addition is currently part of the National Grid Gas plc legal entity.

Due to the expected timing of any transaction, this is not an area of significant risk for our 2015/16 audit, but it has had a major impact on the resource and timing of our audit.

Change in level of risk year on year: New

Although there are no significant accounting impacts in 2015/16 as a result of the transaction process, we have reassessed our risks and materiality benchmarks for UKGD and have worked with management to plan for a significantly accelerated UK component audit timetable.

During the year National Grid has recognised revenue of £15.1bn; £14.2bn of which is mostly related to the regulated segments in the UK and US.

In the UK, National Grid’s revenue is derived from a number of price controls imposed by the UK regulator, Ofgem, when combined with the application of IFRS, revenue recognition involves limited judgement. The majority of revenue is derived from a small number of customers who settle within agreed terms.

In the US, different services and locations are regulated by different authorities and are subject to numerous price controls. Unlike the UK, revenue is earned through the transportation and supply of gas and electricity to end customers, which does involve judgement as a result of the estimate of accrued income for services delivered but not yet billed to these customers. This is determined using a long-established methodology within the Group.

As such revenue recognition is not an area of significant risk for our audit but does require significant time and resource to audit due to the magnitude.

Change in level of risk year on year: No change

In the UK, we have tested the design and operating effectiveness of key controls in relation to the recognition of revenue, with particular focus on controls over the setting of prices compared to those allowed by the Ofgem price controls and we found no material issues that would impact our audit approach.

We have tested the revenue recognised to amounts invoiced to customers and the subsequent receipt of payment from those customers, with no material exceptions noted.

In the US, in respect of transmission and other non-utility revenues, we selected and tested individual transactions to ensure they were appropriately recorded as revenue in the correct period. We inspected proof of cash payments or confirmed amounts with customers where it was possible to do so. We also inspected regulator-approved tariffs to test that amounts charged were consistent with such tariffs.

We found no material issues arising from our work.

For utility revenues, we selected samples of rate classes to test that customer rates were properly updated in the billing systems, and that rate types were assigned to customers consistent with the type of customer and (where appropriate) the volume of usage. We also selected samples of customer bills and tested that such bills were paid by customers and were consistent with the regulator-approved rate plans. For those bills selected that were outstanding at the end of the year, we confirmed the balance with customers, and tested amounts to subsequent cash receipts where no confirmation was received.

With respect to unbilled revenue we tested management’s assumptions in relation to consumption by reference to historical data as well as specific current year factors, including weather patterns. In so doing, we did not note any significant issues which would impact the Group financial statements.

Significant estimates are made in valuing the group’s retirement benefit surplus. Small changes in assumptions and estimates used to value the group’s pension obligation (before deducting scheme assets) would have a significant effect on the group’s financial position.

Our audit procedures included:

• challenging the key assumptions supporting the group’s retirement benefit obligations valuation with input from our own actuarial specialists, including comparing the discount rate, inflation rate, salary, pension increase rates and life expectancy assumptions used against externally derived data; And
• assessing the group’s disclosure in respect of the sensitivity of the surplus to changes in the key assumptions.

The group has significant derivative financial instruments, the valuation of which is determined through the application of valuation techniques which often involve the exercise of judgement and the use of assumptions and estimates. Due to the significance of financial instruments and the related estimation uncertainty, there is a risk that the related financial assets and liabilities are misstated.

Our audit procedures included:

• assessing controls over the identification, measurement and management of derivative financial instruments;
• evaluating the methodologies, inputs and assumptions used by the group in determining fair values, with the help of our own valuation specialist;
• challenging the observable inputs into valuation models, such as quoted prices, by reference to externally available market data to assess whether appropriate inputs are used in the valuation;
• comparing valuations derived from our internal valuation model for a sample of instruments to the fair values determined by the group;
• considering the adequacy of the group’s disclosures about the valuation basis and inputs used in the fair value measurement; and
• assessing whether the financial statement disclosures of fair value risks and sensitivities appropriately reflect the group’s exposure to valuation risk.

Revenue recognition and provision for customer debts are key areas of judgement, particularly in relation to:

• the estimate of the revenue value of water supplied to metered customers between the last meter reading and the period end; and
• identifying properties where there is little prospect cash will be received for revenue that has been billed due to either the occupier not being able to be identified or a past history of nonpayment of bills relating to that property; and
• assessing the recoverability of trade debtors as a proportion of customers do not or are unable to pay their bills.

Our audit procedures included:

• assessing whether appropriate revenue recognition policies are applied through comparison with relevant accounting standards and industry practice, including the policy of not recognising revenue where it is not probable that cash will be received;
• testing the group’s controls over revenue recognition and provision for customer debts, including reconciliations between sales and cash receipts systems and the general ledger;
• assessing the assumptions used to calculate the metered accrued income by ensuring inputs to the calculation have been derived appropriately and recalculating the accrued income with the support of our own modelling specialists;
• assessing the appropriateness of the customer debt provisioning policy based on historical cash collections, credits, re-bills and write off information; and

• assessing the adequacy of the group’s disclosures of its revenue recognition and customer debt provisioning policies, including the estimation uncertainty involved in recording revenue and the bad debt provision.

The group has a substantial capital programme which has been agreed with the Water Services Regulation Authority (Ofwat) and therefore incurs significant annual expenditure in relation to the development and maintenance of both infrastructure and non-infrastructure assets. Expenditure in relation to increasing the capacity or enhancing the network is treated as capital expenditure. Expenditure incurred in maintaining the operating capability of the network is expensed in the year in which it is incurred. Capital projects often contain a combination of enhancement and maintenance activity which are not distinct and therefore the allocation of costs between capital and operating expenditure is inherently judgemental. The costs capitalized include an allocation of overhead costs, relating to the proportion of time spent by support function staff, which is also inherently judgemental and could lead to over capitalisation of expenses.

Our audit procedures included:

• assessing the group’s capitalisation policy for compliance with relevant accounting standards;
• testing controls over the application of the policy to spend incurred in the period including attending capital approval meetings to observe the judgements made and evaluating the documented final conclusions;
• critically assessing the costs capitalised for a sample of projects against the capitalisation policy, focusing on those where actual costs differed significantly to budget;

• agreeing overhead costs incurred to supporting documentation on a sample basis and performed comparative analysis of overheads absorbed into capital projects by category;
• testing a sample of capital accruals to assess the existence of the costs being capitalised through review of ageing of specific project accruals; and
• assessing the adequacy of the group’s disclosures of its capitalisation policy and other related disclosures.

There is uncertainty around estimating and valuing the remaining outstanding business economic loss claims. The determination of the liability is subject to judgement as to the amount that each remaining claim will be settled at.

For the liabilities and contingent liabilities related to the Gulf of Mexico oil spill the primary audit engagement team performed the following audit procedures.

• We walked through and tested the controls designed and operated by the group relating to the provisions and payables for the Gulf of Mexico oil spill.

• We met with the group’s legal team to understand developments across key remaining Gulf of Mexico oil spill matters and their status. We discussed legal developments with the group’s external lawyers, reviewed audit enquiry response letters from external legal counsel and read determinations and judgments made by the courts.

• In respect of the provision for the outstanding business economic loss claims:

• We compared the key assumptions that have been used in the determination of the year end provision, to historical experience. These assumptions are the proportion of outstanding claims that will receive a monetary award and the average monetary claim value.
• We reconciled the number of undetermined claims to third party claims management data.
• We performed sensitivity analyses over average cost per claim assumptions and assessed the potential effect on the provision.

• With regard to Plaintiffs’ Steering Committee settlements, we engaged EY actuarial experts to consider the analysis of available claims data undertaken by management. We corroborated the data used in respect of all claim categories, with specific regard to business economic loss, this being the most complex to estimate. Our testing included understanding and verifying trends in the claim models, considering the approach in respect of all claim categories compared with prior periods.

• We assessed the remaining economic loss and property damage claims from individuals and businesses that either opted out of the PSC settlement and/or were excluded from that settlement. We validated a sample of claims to third party data, assessing the year end closing provision was appropriate.

• We considered the accounting treatment of the liabilities, contingent liabilities and disclosures under
IFRS criteria, to conclude whether these were appropriate in all the circumstances.

We are satisfied that management is able to determine a reliable estimate for the remaining business economic loss claims and that the disclosures presented in relation to contingent liabilities are appropriate.

Based on our procedures we are satisfied that the amounts provided in the financial statements, as disclosed in Note 2 of the financial statements, are supported by claims experience.

2016 has been a more uncertain political and macroeconomic environment, influenced by significant political and economic events.

Additionally, OPEC decisions on the production of hydrocarbons have a significant effect on prices in global commodity markets.

These factors combine to create additional uncertainty in relation to key inputs into future cash flow forecasts, which are used to project the recoverability of the group’s tangible and intangible assets.

We extended the scope of our original planned procedures to address the changing risk. This included further use of EY valuation experts in critically assessing and corroborating, to external market data, the revised assumptions used in impairment testing, the most significant of these being future market oil and gas prices and discount rates. We also focused on reserves and resources volumes, as described elsewhere in our report.

In addressing this risk, audit procedures were performed by component teams at each of the group’s 10 Upstream components scoped-in for the audit of asset impairment. In addition, the primary audit engagement team tested certain remaining assets identified at risk of impairment.

• We walked through and tested the controls designed and operated by the group relating to the assessment of the carrying value of tangible assets.

• We examined the methodology used by management to assess the carrying value of tangible assets assigned to cash-generating units, to determine its compliance with accounting standards and consistency of application.

• We assessed the centrally-derived oil and gas price assumptions by reference to external market data and engaged EY valuation experts to critically assess the suitability of the revised assumptions.

• We evaluated estimates of future cash flows and considered whether these were appropriate in light of future price assumptions and cost budgets.

• Together with EY valuation experts we assessed specific inputs to the determination of the discount rate. Such inputs were benchmarked against rates observable in the markets in which the group operates.

• We performed procedures over the completeness of the impairment charges and reversals and exploration write-offs, also validating that base data used in the impairment models agreed to the underlying books and records.

• We checked the mathematical accuracy of the impairment models.

BP’s oil and gas price assumptions are comparable to the range seen within the industry at this time. The methodology used has been applied consistently year on year.

The pre-tax discount rate of 9% (2015 11%) and the post-tax discount rate of 6% (2015 7%) are within our range of expectation. The movement year on year is verifiable and reasonable based on market data. Based on our procedures, we believe the impairment charges and certain reversals of previous impairment charges recorded, are appropriate and in accordance with IFRS.

Based on our procedures on the exploration portfolio we consider the write-offs were properly recorded and remaining carrying values are appropriate and in accordance with IFRS.

The estimation of oil and natural gas reserves and resources is a significant area of judgement due to the technical uncertainty in assessing quantities and complex contractual arrangements dictating the group’s share of reportable volumes.

Reserves and resources are also a fundamental indicator of the future potential of the group’s performance.

Audit procedures were performed by team members with significant experience of auditing oil and gas reserves, including the primary audit engagement team and component teams at 10 Upstream components.

• We tested the group’s controls over their internal certification process for internal technical and commercial experts who are responsible for reserves and resources estimation.

• We assessed the competence and objectivity of the group’s internal and external experts, to satisfy ourselves they were appropriately qualified to carry out the volumes estimation.

• We performed procedures to assess the reliability of data provided to external experts.

• We confirmed that significant changes in reserves and resources were made in the appropriate period, and in compliance with the Discovered Resources Management Policy (‘DRM-P’).

• Where volumetric movements had a material impact on the financial statements, we validated these volumes against underlying information and documentation as required by the DRM-P, along with checking that assumptions used to estimate reserves and resources were made in compliance with relevant regulations.

• We validated that the updated reserves and resources estimates were included appropriately in the group’s consideration of impairment and in accounting for DD&A.

Based on our procedures we consider that the reserves estimations are reasonable for use in the impairment testing and calculation of DD&A.

Unauthorized trading activity is a fraud risk associated with a potential deliberate misstatement of the group’s trading positions or mis-marking of positions with an intention to:

• Minimize trading losses.
• Maximize trading profits.
• Understate profits or move profits to subsequent periods when bonus ceilings have already been reached, to maximize individual bonuses across financial years.

These acts would lead to a misstatement of the group’s revenue and profits.

Audit procedures on revenue and trading were performed by component teams and the primary engagement team at 7 components across the UK, US and Singapore.

• We walked through and tested the controls designed
and operated by the group over unauthorized trading
activity.

• We identified trades with the highest risk of unauthorized activity so as to focus our testing on these trades.

• We performed existence and completeness testing by confirming a sample of trades with third parties.

• We verified the fair value of a sample of derivatives using contract and external market prices.

• We tested the completeness of the amounts recorded in the financial statements through performing procedures to detect unrecorded liabilities as well as detailed cut-off procedures around sales, purchases, trade receivables and trade payables.

Based on our procedures we identified no matters to report to the Audit Committee.