operating cost special factors – contents - scottish water us/files/second draft business... ·...

Scottish Water: Second Draft Business Plan. Operating Cost Special Factors – Contents

CONTENTS

Contents

1. Operating Cost Special Factors’ Claims ...........................................3

2. Population Sparsity: Determining a Proxy for England and Wales within

Scotland ................................................................................5

3. Bad Debt .............................................................................. 10

4. Cryptosporidium..................................................................... 16

5. Electricity: Higher Unit Electricity Prices at Smaller Rural Sites........... 23

6. Public Septic Tanks ................................................................. 29

7. Sewer Laterals ....................................................................... 31

8. Supply of Chemicals ................................................................ 40

9. Travel ................................................................................. 44

10. Complex Water Treatment ........................................................ 59

11. Household Metering................................................................. 65

Scottish Water: Second Draft Business Plan. Operating Cost Special Factors - Summary

3

EXECUTIVE SUMMARY

1. Operating Cost Special Factors’ Claims

Executive Summary

This document sets out Scottish Water’s Special Factor submission to inform the Strategic Review of

Charges for the period 2010-14.

Scottish Water operates in a different environment from England and Wales (E&W) for reasons of

population sparsity, geography and legal obligation.

The table below provides the value of Scottish Water’s special factor submission, split by each special

factor. This is used in our efficiency analysis, which is detailed further in Operating Cost Appendix 4.5 of

the Second Draft Business Plan.

Value Split

Special Factors 2007/08 prices

£m

Water

%

Sewerage

%

Bad debt 13.3 48 52

Cryptosporidium 1.6 100 0

Electricity 1.8 69 31

Public septic tanks 1.2 0 100

Sewer laterals 7.2 0 100

Supply of chemicals 0.3 100 0

Travel 9.1 50 50

Total special factors 34.5

Scope Adjustments:

Complex water treatment 0.0 100 0

Household metering -6.3 51 49

Total scope adjustments -6.3

Net special factors 28.2

Introduction

This document sets out Scottish Water’s Special Factors submission in relation to the Strategic Review of

Charges 2010-14.

Scottish Water has a different operating environment from E&W for reasons of population density,

geography and legal obligation. We have therefore determined a proxy region for comparison purposes

where information from E&W is not available. This analysis is set out in advance of our detailed special

factors submission and has been used specifically to quantify our travel and electricity special factors.

Scope Adjustments

In the Strategic Review of Charges 2006-10, the Water Industry Commission for Scotland introduced

adjustments to the assessment of Scottish Water’s efficiency position based on their assessment of the

differences in the scope of activities and levels of service provided to customers in Scotland compared to

Scottish Water: Second Draft Business Plan. Operating Cost Special Factors - Summary

4

EXECUTIVE SUMMARY

those in E&W. The Commission used Yorkshire Water as the comparator company for its assessment of

the difference in costs that result from differences in scope of activities.

The value of the scope adjustments could vary significantly depending on which company is at the

frontier in 2007/08. Therefore, we propose that where a scope adjustment is appropriate, the valuation

is based on a comparison to the average company in E&W. This approach is in line with that taken for

special factors.

We present cases covering scope adjustments for household metering and complex water treatment.

Scottish Water: Second Draft Business Plan. Population Sparsity

5

POPULATION SPARSITY

2. Population Sparsity: Determining a Proxy for England and Wales within Scotland

Executive summary

The population density of Scotland is significantly lower than that for England and Wales (E&W) (59 and

355 people per square km, respectively). Scottish Water’s geographical coverage extends across the

most rural regions in the whole of Britain and its operating area is almost three times the size of Anglian

Water, the largest company area in E&W. Scottish Water also provides services to 58 inhabited islands,

unlike the water and sewerage companies (WaSCs) in E&W. As a consequence, Scottish Water has a

significantly different operating environment to the WaSCs1:

• The urban population density of Scottish Water is lower than all of the WaSCs with the

exception of Dŵr Cymru.

• The rural population density is lower than any of the WaSCs and less than half that of

Northumbrian Water, the WaSC with the lowest rural population density.

• The overall population density of Scottish Water is lower than any of the WaSCs and less

than half that of Dŵr Cymru, the WaSC with the lowest overall population density.

This report, firstly, demonstrates Scottish Water’s ‘super-rurality’ position amongst the water and

sewerage companies in Great Britain. This makes it an outlier in respect of population (or asset) density

explanatory variables used in the econometric benchmarking models.

Secondly, this report provides the arguments to support the choice of our Ayr, Clyde and Tweed regions

as the Scottish proxy region for an English and Welsh WaSC. We need to establish such a proxy region in

order to be able to undertake a comparative analysis of operational costs between this proxy area and

the rest of Scotland where relevant English and Welsh data are not available.

Rurality in the GB context

The greater sparsity of the population served by Scottish Water compared to that of the WaSCs in E&W

gives rise to a significantly different operating environment. The table below shows the spread of

population and asset densities within Scotland and between Scotland and the WaSCs. Only Ayr and

Tweed are close to the lowest-density WaSCs. The Clyde region is more densely populated than Thames

Water and the remaining regions are significantly less densely populated than any WaSC in E&W.

1 Second Draft Special Factors Submission, April 2005. We have since reviewed the recent mid-2006 population estimates (National Statistics) to verify that the recent increase in migration to the UK has not changed the picture significantly. It shows that E&W have become even more densely populated than Scotland.


6

POPULATION SPARSITY Scottish Water Operational Regions and WaSCs – density comparisons

Wastewater Water

Company / Region

Population

Density

(Sewerage

population

per square km

of land area)

Sewerage

Service

Population per

wastewater

treatment

works2

Sewerage

Service

Population per

km length of

Sewer

Water Service

Population per

water

treatment

works

Water Service

Population per

km length of

main

Ayr 153 6,714 97 43,631 135

Clyde 1,439 68,818 121 110,073 227

Don 42 1,701 82 11,683 66

Forth 43 2,170 92 9,114 103

Ness 6 304 55 1,721 35

Nith 82 2,167 95 30,731 97

Tay 73 2,941 93 40,781 109

Tweed 134 5,828 99 37,487 152

Proxy (Ayr, Clyde, Tweed)

204 8,936 105 50,690 165

Scottish Water 59 2,511 95 15,906 106

Anglian 210 5,173 133 30,072 114

Dwr Cymru 143 3,748 165 40,032 105

Northumbrian 290 6,369 169 103,183 230

Severn Trent 399 8,140 157 56,299 161

South West 148 2,439 175 40,717 108

Southern 405 11,548 199 23,244 170

Thames 1,023 39,086 200 89,937 272

United Utilities 487 11,783 173 62,433 162

Wessex 263 6,477 159 11,334 109

Yorkshire 359 7,768 161 71,594 155

Average (E&W) 355 8,432 170 49,681 163

Note: Data taken from Ofwat 2007/08 Annual Returns and Scottish Water’s 2007/08 Annual Return

2 Includes septic tanks


7

POPULATION SPARSITY

ANG

WSH

SVT

SWT

TMS

UUT

NES

SRN

WSX

YKY

Scottish Water

Average

0

2

4

6

8

10

12

14

0 10 20 30 40 50 60 70 80 90

Area (1000 km2)

Po

pu

lati

on

(m

illio

ns

)Population Density Comparison of all WaSCs and Scottish Water

The chart above illustrates population density for all WaSCs. It shows that Scottish Water’s population

density is an outlier when compared with the WaSCs because its population per geographical area is far

lower.

Determination of a Scottish Water proxy area for England and Wales

Looking at the different density measures shown above, it can be seen that none of Scottish Water’s

eight regions are directly comparable to any of the WaSCs. We have used population density as the main

parameter to determine rurality since it is entirely outside Scottish Water’s control. Several alternative

proxy areas were evaluated. From this analysis, the combination of Ayr, Clyde and Tweed regions is

considered the most appropriate proxy for the average WaSC in E&W, since its population density falls

within the range of those of the E&W WaSCs.

Ayr, Clyde and Tweed have been grouped together to form a comparator area within Scottish Water. The

other five Scottish Water regions are significantly less densely populated than the WaSC areas in E&W

and are deemed to be causing the extra, rurality-related costs where they are not appropriately captured

in the econometric models. Even then, the proxy chosen is more rural than an average company in E&W;

when applied, for example, to the travel special factor, it will generate a conservatively low estimate of

additional costs.


8

POPULATION SPARSITY Annex 1: Abbreviations

Company Abbreviation Water & Sewerage Company (WaSC) / Scottish Water operational region

Scottish Water - Scottish Water

Ayr Region Ayr Ayr Region – Scottish Water

Clyde Region Clyde Clyde Region - Scottish Water

Don Region Don Don Region - Scottish Water

Forth Region Forth Forth Region - Scottish Water

Ness Region Ness Ness Region - Scottish Water

Nith Region Nith Nith Region - Scottish Water

Tay Region Tay Tay Region - Scottish Water

Tweed Region Tweed Tweed Region - Scottish Water

Anglian Water ANG WaSC

Dwr Cymru WSH WaSC

Northumbrian Water NWT WaSC

Severn Trent Water SVT WaSC

South West Water SWT WaSC

Southern Water SRN WaSC

Thames Water TMS WaSC

United Utilities UUT WaSC

Wessex Water WSX WaSC

Yorkshire Water YKY WaSC


9

POPULATION SPARSITY

Annex 2: Map of Scottish Water Operational Regions

Scottish Water: Second Draft Business Plan. Bad Debt

10

BAD DEBT

3. Bad Debt

Executive Summary

Scottish Water is currently required by the Scottish Government to use the local authorities to bill,

collect and recover household water and waste water charges.

The level of bad debt incurred in Scotland is higher than experienced with England and Wales (E&W).

This extra cost is out-with managerial control and as such requires a special factor adjustment.

The additional operating costs for Scottish Water’s high level of household bad debt is valued at £13.3m

in 2007/08 prices. This is based on Scottish Water’s reported underlying level of household bad debt

being 4.39% of net charges billed in 2007/08.

Basis for the Special Factor

Background and Management Control

In 1996, when the responsibility for the provision of water and waste services moved to the water

authorities from local government, the former Scottish Office required that councils continued to be the

mechanism for billing and collecting household revenues for water and waste water charges on behalf of

the water authorities. The “Paying for Water Services 2006-2010” consultation of July 2004 stated that

this arrangement will continue until 2010 at the earliest.

The situation is different in E&W where the water and sewerage companies (WaSCs) are directly

responsible for billing, collecting and recovering their household charges. They are therefore more able

to control their collection and bad debt levels.

Despite Scottish councils having improved their collection performance in recent years (refer to Annex

3), poor levels of initial collection contributed to Scottish Water’s level of bad debt remaining higher

than the E&W average.

Quantification

Methodology

Scottish Water carried out an anonymised benchmarking exercise with the WaSCs. Each company was

asked to provide information on their bad debt split by households and non-households.

To determine the additional operating costs associated with Scottish Water’s household billing and

collection, the following approach was adopted:

• Establish what Scottish Water’s household bad debt costs and household billing and

collection costs would be if the average levels experienced in E&W applied in Scotland.

• Quantify the difference between Scottish Water’s actual household bad debt costs and those

using the E&W average.

This is encapsulated in the following formula:


11

BAD DEBT

Household bad debt special factor

=

Scottish Water cost of bad debt – SW cost of bad debt using E&W average

+

Scottish Water cost of collection – Scottish Water cost of collection using E&W average

Scottish Water – Bad debt

The table below shows the average household bad debt per property in Scotland in 2007/08 prices.

Company Household Bad Debt

(000s) No. of Household Properties (000s)

Average level of household bad debt per

property

Scottish Water £29,590 4,550 £6.50

Note: A household with both water and sewerage is counted as two properties.

The table above reflects Scottish Water’s underlying level of bad debt rather than that reported in the

2007/08 annual return. Scottish Water’s underlying level of bad debt was assessed by adjusting the

annual return 2007/08 charge for the effective release of prior year over provisions. The underlying level

of bad debt was estimated to be 4.39% of net billed charges. Refer to Operating Cost Appendix 4.1 of

the second draft business plan.

England and Wales – Bad debt

The level of debt in E&W is provided by each WaSC in their June Returns as a combined household and

non-household figure. Each of the WaSCs have been approached for benchmarking analysis and asked to

identify the level of household bad debt. Eight of the WaSCs, supplying 82% of households, provided this

breakdown. The average household bad debt per property in E&W has been taken as the weighted

average of the benchmarked companies. The average E&W household debt is £3.19 per property in

2007/08 prices as shown below.


12

BAD DEBT

Average household bad debt per household property England and Wales

Company Household Bad Debt

(£000s) No. of Household Properties (000s)*

Average level of household debt per

property

Company 1 Benchmarked Confidential Confidential Confidential




Company 5 Benchmarked Confidential Confidential Confidential**

Company 6 Benchmarked Confidential Confidential Confidential***

Company 7 Benchmarked Confidential Confidential Confidential***


Average E&W £100,823 31,616 £3.19

*A household with both water and sewerage is counted as two properties.

** For these companies, debt per property is calculated using 2006/07 bad debt but inflated to 2007/08

prices divided by the number of properties in 2007/08.

*** 2007/08 Benchmarked figures for household and non-household bad debt adjusted to the total bad

debt figure as reported in June Return 2008.

The order of companies in this table and the corresponding table for cost of collection are not the same.

Quantifying the Bad Debt Special Factor

If the average level of household bad debt per property in E&W is applied to the number of Scottish

Water households the bad debt factor can be quantified as follows:

• Scottish Water household bad debt £29.6m

• Less E&W average bad debt x number SW properties £14.5m [£3.19 x 4.55m]

• Value of bad debt special factor £15.1m

If Scottish Water was able to achieve the average bad debt performance in E&W the cost of bad debt

would reduce by £15.1m in 2007/08 prices.

Cost of the Billing and Collection Service

Scottish Water is obliged to use the councils for billing and collection. As the councils provide this

service at a lower cost to that experienced in E&W this is used as an offsetting factor.

The table below shows the cost to Scottish Water of collecting household charges in 2007/08 prices.

These costs are sourced directly from Scottish Water’s General Ledger.

Household cost of collection per property

Company Cost of Collection – Household (£000s)

No. of Household Properties (000s)

Cost of Collection per Cost per property

Scottish Water 14,316 4,550 £3.15


13

BAD DEBT

The household billing and collection data shown below for E&W have been sourced from benchmarking

analysis. Each company was asked to provide their cost of collection, billing and debt recovery.

Information was provided by seven companies, supplying 69% of households. The weighted average cost

of the seven reporting companies is shown in the following table.

Household cost of collection per property in England and Wales

Company Cost of Collection – Household (£000s)

No. of Household Properties (000s)*

Cost of Collection per Cost per property**








Average E&W £93,561 26,432 £3.54

Note: The order of companies in this table and the corresponding table for bad debt are not the same

*A household with both water and sewerage is counted as two properties.

** For these companies, cost of collection is calculated using 2006/07 cost of collection inflated to

2007/08 prices then divided by the 2007/08 number of properties. The 2006/07 figures were £2.22 and

£2.28 per property respectively.

Quantifying the Collection Factor

If the average cost of collection per property in E&W is applied to the number of Scottish Water

households the cost of collection factor can be quantified as follows:

• Scottish Water household cost of collection £14.3m

• Less E&W average cost of collection x number SW properties £16.1m [£3.54 x 4.55m]

• Value of collection factor -£1.8m

If Scottish Water experienced the same collection levels as those experienced in E&W the cost of

collection would increase by £1.8m in 2007/08 prices.

Mitigating the Impact

Scottish Water is focusing on mitigating any negative effect associated with the obligation to use the

councils. We do this by incentivising the councils to improve their collection rates.

Scottish Water has reviewed the Service Level Agreements (SLAs) with all 32 councils. In 2004/5 there

were 9 SLAs with collection incentives representing 31.4% of households. There are now 27 SLAs with

collection incentives covering 88.5% of households.


14

BAD DEBT

Collection incentives built into the SLAs for 2007/08 are as follows:

• 17 councils have a current year collection incentive

• 10 councils have a current and prior year incentive

• 2 councils’ performance incentives are currently under negotiation

The remaining councils have not agreed to have collection incentives built into their SLAs, however it

should be noted that they have higher than average collection rates.

In-year billing collection statistics have improved from 91.85% in 2004/05 to 93.44% in 2007/08 (see

Annex 3). Local authorities also have their own initiatives to improve Council Tax collection rates, which

in turn have an impact on the collection of Water and Sewerage charges.

Conclusion

The additional operating costs incurred relative to E&W is valued at £13.3m in 2007/08 prices. This is

based on Scottish Water’s underlying level of household bad debt being 4.39% of net charges billed in

2007/08.


15

BAD DEBT

Annex 3: Household In-Year Collection* - CONFIDENTIAL

Scottish Water: Second Draft Business Plan. Cryptosporidium

16

CRYPTOSPORIDIUM

4. Cryptosporidium

Executive Summary

The Cryptosporidium (Scottish Water) Directions 2003 place obligations on Scottish Water to carry out

analysis on raw water and treated water at each water treatment works (WTW), at frequencies

determined by the overall risk assessment score associated with that works. This sampling requirement

is in excess of sampling programmes undertaken by the companies in England and Wales (E&W). The

significant differences between the legislation in Scotland and in E&W are:

• The Directions require Scottish Water to sample raw water supplies. This is not required in

E&W.

• The Directions require Scottish Water to sample treated water at all WTWs. Water

companies in E&W are only required to sample at WTWs considered at ‘significant risk’.

Scottish Water samples more WTWs for cryptosporidium than the whole of E&W.

The Scottish Water claim is based on the number of treated water samples taken at low and moderate

risk WTWs and the number of raw water samples that would not be taken if Scottish Water were subject

to E&W legislation. The excess number of samples taken in Scotland is 13,809 based on this approach.

The value of the special factor claim is £1.6m in 2007/08 prices.



The Drinking Water Directive was transposed into appropriate regulations in the UK. In E&W, The Water

Supply (Water Quality) Regulations 1989 applied and in Scotland, The Water Supply (Water Quality)

(Scotland) Regulations 1990 applied. Both sets of regulations were superseded in December 2003 by The

Water Supply (Water Quality) Regulations 2000 in E&W and The Water Supply (Water Quality) (Scotland)

Regulations 2001 in Scotland.

In general terms both sets of regulations are identical. There is, however, a significant difference in how

these regulations are implemented north and south of the border. This results in the sampling

requirement for Scottish Water being in excess of the sampling programmes carried out by the companies

in E&W.

Sampling Requirements in England and Wales

In E&W the Drinking Water Inspectorate (DWI) took the view that an amendment to the regulations was

required to take account of the increasing risk of cryptosporidium. To this end it incorporated a

treatment standard for cryptosporidium of 1 oocsyt per 10 litres of treated water. Within the

regulations, water companies are obliged to carry out risk assessments to determine whether, in their

judgement, a works carries a significant risk of a breach of this standard. If there is a significant risk the

water company is obliged to carry out the necessary measures to ensure compliance or implement a

suitable sampling regime to monitor for cryptosporidium. Only 11% of the WTWs in E&W are categorised

as at significant risk. Where the risk assessment does not identify a ‘significant’ likelihood of breaching


17

CRYPTOSPORIDIUM

the cryptosporidium standard the E&W companies are under no obligation to sample for cryptosporidium.

This standard applies for the base year 2007/08.

An amendment to the 2000 Regulations applies from 31 December 2007 which removes the treatment

standard for cryptosporidium. This means that a similar situation applies as in Scotland i.e. that

cryptosporidium is covered by the wholesomeness catchall (see below) and that companies are liable to

prosecution if any level of cryptosporidium enters supply. However, no sampling regime has been

prescribed to companies in E&W as companies can choose how best to manage and monitor their risk.

Sampling Requirements in Scotland

In Scotland, the Scottish Executive (now Scottish Government) took the view that the existing regulations

already covered cryptosporidium by way of the wholesomeness catchall (i.e. the water does not contain

any element, organism or substance at a concentration or value which would be detrimental to public

health). To complement the regulations the Scottish Executive issued The Cryptosporidium Direction

2000 obliging the water authorities to carry out risk assessments on all works and carry out remedial

works to ensure that all water treatment works carried a moderate to low risk. The Direction also gave a

benchmark standard that all authorities had to adhere to in terms of monitoring, sampling and analysis

and operational best practice. This was based on Badenoch reports. Later Directions were based on the

Bouchier report.

The Cryptosporidium (Scottish Water) Directions 2003 came into force on 31 December 2003. This set

the baseline levels of cryptosporidium sampling and analysis required (raw and treated water) for each

supply. Details of the sampling requirements in Scotland can be seen in Annex 4.

Sampling Taken in Scotland, England and Wales, 2007

The following table compares numbers of cryptosporidium samples taken at water treatment works in

Great Britain. This shows that Scottish Water has to monitor over twice as many water treatment works

as all the companies in E&W combined.

Cryptosporidium Sampling in England, Wales & Scotland in 2007

E&W Region Number of WTWs Monitored WTWs % WTWs monitored

Number of Samples taken

Northern 232 39 16.8 11,845

Eastern 191 9 4.7 3,133

Midlands 212 11 5.2 3,960

Southern 227 22 9.7 5,905

Thames 195 17 8.7 6,178

Western 153 31 20.3 9,681

Wales 96 20 20.8 6,953

Total E&W 1,306 149 11.4 47,654

Scotland 307 307 100 16,742

Notes: (i) E&W regions data from DWI Drinking Water Quality Report 2007.

(ii) Scottish Water sampling numbers from LIMS. WTW numbers from Water Quality Report 2007

(iii) E&W samples WTWs of significant risk. Scottish Water samples all works regardless of risk.


18

CRYPTOSPORIDIUM

As Scottish Water must comply with the Cryptosporidium (Scottish Water) Direction 2003, the additional

samples required under this Direction are outside management control.

In PR99, companies in E&W applied for, and were granted by Ofwat, investment funding to install robust

treatment to meet the Cryptosporidium Standard. This explains why just 11% of WTWs are at significant

risk and hence reduces the sampling requirements. In Scotland, The Cryptosporidium (Scottish Water)

Directions 2003 required Scottish Water to establish an extensive sampling programme to build a body of

data to understand the extent of cryptosporidium risk exposure. As a result of this monitoring,

investment to reduce cryptosporidium in water supplies is being included to meet a Ministerial objective

for the first time in the 2010-14 period.

Econometric Models

Sampling and analysis costs are captured in the Water Business Activities model. Variations in sampling

requirements for cryptosporidium are not modelled. Predicted opex in this model is based upon the

number of billed properties served. Therefore not only does the model not take account of variations in

legislation, but also it does not account for variations in the asset base necessary to serve the

population. This special factor methodology reflects the additional costs that arise as a result of the

legislative difference only.

Quantification

The requirement for sampling in E&W is restricted to those WTWs where the cryptosporidium risk is

considered to be significant. The method of determination of risk in E&W is judgmental, based on

knowledge and experience of the WTW and catchments. The method of determination of risk in Scotland

is based on a formula and scoring system, which establishes WTWs at high, moderate or low risk.

Scottish Water takes samples for cryptosporidium at all of its WTW, however, only 25 are classed as high

risk (see Annex 4). The number of treated water samples taken at the high risk WTW is 2,933

The method for calculating the claim is as following:

Scottish Water actual

number of samples taken

16,742

- Number of samples taken for treated water

at Scottish Water high risk WTWs

2933

*

SW cost per

sample

£119

The excess number of samples taken under this method in Scotland is 13,809. At a cost for sampling and

analysis of £119 per sample the value of the special factor claim is £1.6m.

It has been assumed that a high risk site in Scotland is broadly equivalent to a site of significant risk in

E&W and therefore sampling at low and moderate risk sites and raw water sampling are additional to the

sampling requirement in E&W3. This is supported by the statistic that 11% of sites in E&W are regarded

as at significant risk compared with 9% of sites in Scotland at high risk.

3 From DWI report for E&W Regions on Water Quality 2006


19

CRYPTOSPORIDIUM


Cost per Cryptosporidium Sample

The cost of Scottish Water’s sampling and analysis has been determined as £119 per sample in 2007/08

prices (see Annex 5). This £119 per sample compares with the £140 per sample (2002/03 prices) in 2003

and represents a reduction in real terms of 25%. This reflects improvements achieved in sampling and

analysis for Cryptosporidium over the intervening period and an adjustment to the percentage of

overheads allocated to Scientific Services.

In 2008 Scottish Water has sought benchmarking information on an anonymised basis from companies in

E&W. The cost per sample ranges between £91 and £242 for sampling and analysis as shown in the

following table.

Comparison of Cryptosporidium Sampling & Analysis Costs

WaSC No.1*

WaSC No.2*

WaSC No.3*

WaSC No.4**

WaSC No.5**

WaSC No.6**

WaSC No.7**

WaSC No.8**

Cost per sample in E&W

Confidential

*2007/08 prices **2006/07 prices

This suggests that Scottish Water costs are within reasonable E&W range of sampling costs.

Optimising the Sampling Collection Process

In order to minimise the cost of cryptosporidium sampling all necessary samples are taken by a

multifunctional sampling team during a visit to a treatment works and not restricted to cryptosporidium

or other type of sampling. All sampling is regionally coordinated to optimise routes and minimise

distances travelled.

At smaller more remote sites, operations staff will install the sampling filter during routine operational

visits to reduce the number of visits required by the sampling team.

The Cryptosporidium Compliance group has established a new procedure whereby any broken or failing

filter, which results in a missed sample, is reported and the sample is not rescheduled until a report is

returned that the repair is complete. This helps reduce unnecessary trips.

Risk Reduction

Generally Scottish Water does not own catchments supplying WTWs so it is difficult to reduce the risk

from these. On the treatment side, there are a large number of WTWs with robust treatment installed

which, as a by product, reduces the risk of cryptosporidium entering supply, and hence reducing the

required numbed of samples to be taken. Ongoing investment in 2006-10 and planned investment for

2010-14 will further improve treatment and reduce the required number of samples.

Rationalisation

As part of any investment planning process, we seek opportunities to reduce the number of WTWs we

operate by rationalisation of small WTWs and replacement where possible by additional pipelines. This


20

CRYPTOSPORIDIUM

has resulted in the closure of 20 WTWs since 2005/06 and 58 since 2003/04, thus reducing the number of

samples required to be taken and analysed.

Number of Scottish Water’s Water Treatment Works

2003/04 2004/05 2005/06 2006/07 2007/08

Number of water treatment works

371 368 333 319 313

Note: Figures taken from Scottish Water Annual Returns, table E4.

As a result of maintaining old WTW during the commissioning of replacement WTWs, Scottish Water

operated 313 works over the year but only sampled from 307 at any point in time. The rationalisation of

WTWs is an ongoing process.

Conclusion

Scottish Water must undertake significantly more cryptosporidium sampling and analysis than in E&W as a

result of the difference in legislation. A special factor claim to the value of £1.6m in 2007/08 prices is

submitted to reflect the additional operating costs associated with cryptosporidium sampling.


21

CRYPTOSPORIDIUM

Annex 4: Final water and raw water sampling requirements in Scotland

Sampling requirements for final and raw water are given in the following tables.

Final Water Sampling Requirements

WTW Maximum Design Flow (Ml/day)

Final Water ≤ 1 > 1 ≤ 10 > 10 ≤ 50 > 50

High risk

( > 55) 52 104 365 365

Moderate risk (35 - 54)

12 52 52 104

Catchment & Treatment Risk Score

Low risk

(< 35) 12 12 52 52

The risk score is calculated by using a methodology that considers factors that include: the degree of

exposure of the catchment to oocysts, agricultural practices, sewage inputs, water source type, river and

intake management, water treatment, treatment works monitoring, performance and operational

factors, and cryptosporidium monitoring. The methodology is based on a simple scoring system that

assesses the risk by identifying the potential for cryptosporidium to be present in the water4.

In contrast to companies in E&W, Scottish Water is required to sample for cryptosporidium at all WTW

regardless of the level of risk. There is also a requirement in Scotland for sampling of raw water which is

not required in E&W.

Raw Water Sampling Requirements

WTW Maximum Design Flow (Ml/day)

Raw Water ≤ 1 > 1 ≤ 10 > 10 ≤ 50 > 50

High risk

( > 55) 12 26 52 52

Moderate risk (35 - 54)

0 12 12 26

Catchment & Treatment Risk Score

Low risk

(< 35) 0 0 12 12

The determination of whether a WTW is considered to be of a high, moderate or low risk uses a

combination of the risk scores for raw and treated water and applies a weighting according to the

population served. The result of which is 25 of Scottish Water’s WTW are considered high risk.

4 Cryptosporidium Direction 2003, page 16


22

CRYPTOSPORIDIUM

Annex 5: Determination of Sampling and Analysis Cost

• Sampling costs are based on the actual costs for financial year 2007/08 and the actual

number of cryptosporidium samples taken during 2007/08.

• Sampling costs are based on total 2007/08 sampling costs divided by total no of samples with

an adjustment for cryptosporidium to allow two trips for each cryptosporidium sample; one

to put the filter on and one to take it off.

• Sampling costs include for the cost of the filter at £20.20 each.

• Analysis costs are based on actual cryptosporidium analysis costs divided by the number of

cryptosporidium samples.

• Laboratory on-costs from Activity based Management (ABM) are an additional 31.6% and

include for all management, QA, equipment replacement and internal recharge for HR, IT

etc.

Cryptosporidium Sampling and Analysis Costs

Cost per sample

Cost of sampling £51.96

Cost of analysis £38.23

Total direct cost of sampling and analysis £90.19

Overhead uplift for laboratory (from ABM): £28.48 (31.6% of £90.19)

Total cost of sampling and analysis per sample £118.67

Scottish Water: Second Draft Business Plan. Electricity

23

ELECTRICITY

5. Electricity: Higher Unit Electricity Prices at Smaller Rural Sites

Executive summary

Scottish Water incurs costs above those predicted by the econometric model because it needs to operate

a high number of remote, low-power assets. The geography and population sparsity of Scotland,

particularly in the Highlands and Islands, necessitate many small treatment works and pumping stations

to serve small communities. These small sites are charged at a more expensive rate than large sites.

This special factor claim is £1.8m in 2007/08 prices.

Background

Scottish Water’s sites are generally assigned to three broad tariffs depending upon the level of

consumption5:

• Block tariff (profile codes 03 & 04): Typically aimed at small businesses that do not consume

sufficient kWh to benefit from economies of scale. In addition, this tariff is devised for

small use customers who tend to operate at times when electricity is most expensive (i.e.

during the day, Monday to Friday) and not during evenings and weekends. It comprises

simply a daily standing charge and a volumetric charge.

• Maximum demand tariff (profile codes 05-08): For moderate consumers of electricity with a

highest monthly demand of between 45kVa and 100kVa. The tariff is more complex,

comprising a standing charge, a capacity charge, a maximum demand charge and a

volumetric charge which varies by time of day or night.

• Half-hourly (profile code 00): Designed for very large sites whose highest monthly demand

exceeds 100kVA and which have meters that record half-hourly consumption. Each site has a

different tariff structure depending on a number of factors that reflect the variability of

consumption, the maximum demand on the system, the supply voltage and losses specific to

that site relating to the distance from the transmission network.

Scotland includes the most rural and sparsely populated areas in the UK6. Many of Scottish Water’s

customers in the Highlands and Islands live in small remote settlements. Therefore, Scottish Water’s

assets tend to be smaller but more numerous than in E&W. Individually, these sites use relatively little

electricity. As a result, of the total of 4,533 supply points, 3,865 are assigned to the more expensive

Block Tariffs. A larger share of these electricity costs cover standing charges. By contrast, sites that

consume larger volumes of electricity are placed on a Half-Hourly tariff. Scottish Water’s biggest

electricity consuming sites are placed on the Half-Hourly tariff. These 423 supply points account for

approximately 77% of Scottish Water’s electricity consumption.

5 A few legacy contracts remain where supply points are still charged on the domestic tariff, however these amount to very little consumption or cost. Supply points are no longer assigned to this tariff.

6 See ‘Population Sparsity – Determining a Proxy for E&W within Scotland’ report for further detail.


24

ELECTRICITY

Total number of Supply Points 2007/08

Region Tariff

Ayr Clyde Don Forth Ness Nith Tay Tweed

Scottish Water

00 HH 58 23 60 50 48 66 65 53 423

01 & 02 Dom 5 1 9 13 4 14 9 15 70

03 & 04 BT 325 186 594 500 706 463 702 389 3,865

05-08 MD 3 7 44 26 21 18 39 17 175

Total 391 217 707 589 779 561 815 474 4,533

Note: HH: half-hourly, Dom: domestic, BT: block tariff, MD: maximum demand.

A comparison of supply points by tariff with E&W shows the difference in demand. Scottish Water

commissioned an independent report7 to compare electricity prices with water companies in E&W. Five

companies contributed to this study representing a spread of both predominantly urban and rural service

areas and is therefore representative of the UK water industry.

Percentage of Electricity Consumed on Each Tariff 2007/08

Tariff Scottish Water

(%) E&W Average* (%)

00 Half-hourly 76.8 Confidential

01-02 Domestic 0.3 Confidential

03-04 Block Tariff 18.4 Confidential

05-08 Maximum Demand 4.5 Confidential

*EIC analysis based upon [Confidential]

Consumption & Cost of Supply Points by Tariff as % of Regional Total 2007/08

Region Tariff

Ayr Clyde Don Forth Ness Nith Tay Tweed

Scottish Water

00 82% 96% 71% 63% 48% 78% 70% 78% 77%

03-04 17% 4% 21% 28% 44% 19% 21% 18% 18% % Region Consumption

05-08 1% 0% 8% 8% 7% 3% 9% 3% 4%

00 77% 93% 67% 57% 44% 72% 63% 73% 71%

03-04 22% 6% 26% 34% 49% 24% 27% 23% 24% % Region Cost

05-08 1% 0% 7% 9% 6% 4% 10% 4% 5%

Water companies in E&W consume a greater proportion of electricity on cheaper half-hourly rates than

Scottish Water. 71% of Scottish Water’s cost is attributed to the 423 large half-hourly supply points. For

Scottish Water as a whole, electricity consumed on the block tariff amounts to 18% of total consumption

and 24% of all electricity costs.

This demand structure puts additional unavoidable cost pressures on Scottish Water, compared to E&W

water companies. Scottish Water has optimised sites to the best tariff as far as possible.

7 Electricity Benchmarking Analysis for Scottish Water, April 2008. Energy Information Centre


25

ELECTRICITY

Econometric Models

The cost differentials caused by unavoidable asset dispersion are not taken into account in all of the

econometric models. Differences in the split of consumption between the various tariff types is not

reflected in the models, hence the need for Scottish Water to claim for the additional costs it incurs by

having a higher proportion of consumption on the more expensive non-half-hourly tariffs.

On the wastewater side, power costs are included in the total cost figure of each model. However, scale

is only adequately taken into consideration in the sewage treatment works models. Therefore, power

consumed in wastewater treatment has been excluded from this claim; wastewater networks and sludge

power costs are included.

On the water side, there is a separate model for power consumption that relates costs to distribution

input per average pumping head. The model specification links power costs only to this aggregate

variable. This does not take into account the dispersion of production and pumping of water that leads

to higher power costs in Scotland, due to a larger number of low-powered assets on high tariffs. Further

details are provided in Annex 6.

Quantification

If the more rural regions of Scotland had similar population distributions as E&W, these regions would not

require the same number of assets (and, by implication, supply points). To calculate the additional

electricity cost caused by higher consumption on block tariff supply points, the consumption profile of

E&W has been applied to Scottish Water. The difference between the two scenarios shows the impact of

the different demand structure. Electricity consumption for wastewater treatment is excluded from the

calculation tabulated below as the econometric model is a unit cost model and therefore could reflect

scale of electricity consumption.

Applying the average E&W consumption profile to Scottish Water’s tariffs (excluding wastewater

treatment works)

Total MWh excl consumption at wastewater treatment (A)

Tariff E&W % on Tariff

(B) Scottish Water pence/kWh (C)

Cost (A*B*C)

£m

00 (EHV = 33kV) Confidential 6.00 Confidential

00 (HV = 11kV) Confidential 6.39 Confidential

00 (LV = 0.4kV) Confidential 7.41 Confidential

01 Confidential 8.91 Confidential




375,270

05-08 Confidential 7.95 Confidential

Total 26.44

The table above demonstrates that Scottish Water’s total electricity costs excluding wastewater

treatment would be £26.4 million when the average E&W split between tariffs is applied. Scottish


26

ELECTRICITY

Water’s actual cost excluding wastewater treatment is £28.2 million. This amounts to a difference of

£1.8 million in 2007/08 prices.

Mitigation

Scottish Water’s relatively low water pumping head demonstrates that it has been taking advantage of

the opportunities of gravity-feeding the water supply. The scope for an amalgamation of assets is limited

by;

• The water resources available

• The fact that the additional network expenditure necessary would far outweigh the power

cost disadvantage

Benefit realisation of moving to Half-Hourly Tariffs

The Meter Expansion Programme was completed in September 2006. The aim of this was to identify

further sites with a load factor of greater than 40% that could be placed on the half-hourly tariff.

Benefit realisation within Non-Half-Hourly Tariffs

A review of tariffs at 320 operational has been agreed with the supplier, delivering a forecast saving of

around £120,000/annum.

Consumption Reduction Initiatives

All buildings undergoing refurbishment are fitted with Passive Infra Red (PIR) motion sensors to minimise

electricity used in lighting. Currently 13 out of 51 buildings (Offices, labs, depots) already have PIR

sensors fitted and these include all major buildings/complexes.

Major buildings have Building Management Systems to control the temperature and the timing of central

heating.

Limitations to potential for mitigating electricity costs

Distribution and transmission use of system charges are regulated by Ofgem. They cannot be avoided nor

negotiated. Distribution charges are higher on average in Scotland as these reflect the increased cost to

the electricity provider of maintaining a network in sparsely populated areas. Therefore, Scottish Water

cannot take action to mitigate the costs of these higher charges. For details on the cost of delivery see

Annex 7.

Conclusion

Scottish Water incurs costs above those predicted by the econometric model because it operates a high

number of remote, low-power assets. The special factors claim is £1.8m in 2007/08 prices.


27

ELECTRICITY

Annex 6: Econometric modelling of water power costs

For water, power costs are modelled separately from all other operational costs. Within the model used,

there is no measure for the total number and the sizes of individual sites. The explanatory variable in

the model is the product of average pumping head and total distribution input.

The Commission has indicated that it regards a possible ‘economies of scale’ effect in the model as

sufficient to account for the higher costs incurred by Scottish Water. We do not agree with this because

the unit cost reduction implicit in an increase in the explanatory variable cannot be interpreted as an

‘economies of scale’ effect. The reasons for this are outlined below.

Although there is a small reduction in estimated costs per Ml per m pumped as the explanatory variable

increases, the value of the explanatory variable rises if either average pumping head or distribution input

increases, leading to counter-intuitive results.

Scenario (i) – current position (ii) – double pumping head

(iii) – double distribution input

Pumping head 56m 112m 56m

Distribution input 2,271 Ml/d 2,271 Ml/d 4,542Ml/d

Predicted cost £13,042,000 £24,534,000 £24,534,000

Unit cost per Ml/d per m head £103 £96 £96

For example, if Scottish Water were to run services for an additional identical area, distribution input

would double but average pumping head remain constant. This is because changes in the numerator and

denominator to calculate average pumping head would each double, leaving average pumping head to be

the same. As a result of increased distribution input, the model would inappropriately allocate a

reduced unit power cost. Essentially, the model would interpret this as a ‘scale’ effect even though this

does not reflect the realities of such a situation, which are that unit costs should remain the same and

total costs should rise accordingly.

The central point is that this special factor claim contends that the geography of Scottish Water’s area

means that Scottish Water incurs higher than predicted power costs due to its many very remote assets

that can only be connected to the grid at relatively disadvantageous rates. A model which is based on

overall volume pumped to a certain height and estimated using the data from companies that face

fundamentally different geographical situations is not able to reflect adequately Scottish Water’s extra

costs that are due to the wide dispersion of its customers.


28

ELECTRICITY

Annex 7: Delivery Charges

Delivery charges are the component of the total electricity cost that is regulated by Ofgem and therefore

outside managerial control. They cannot be avoided nor negotiated. EIC has identified that: “The net

effect is that Scottish Water is currently paying an average of 0.12p/kWh more for its delivery charges

than if the average English and Welsh indices applied. This equates to an annual sum of £526,747.”

The table below compares delivery charges in Scotland with E&W using an index to maintain

confidentiality of data.

Assessed delivery prices index

SW E&W

HH HV 100 Confidential

HH LV 100 Confidential

Profile 03 100 Confidential

Profile 04 100 Confidential

Profiles 05-08 100 Confidential

Weighted Average 100 82

Note: EIC analysis based upon [confidential]

We are not adding this onto the calculated claim of £1.8m because we are of the view that it is largely

included in the previous calculations.

Scottish Water: Second Draft Business Plan. Public Septic Tanks

29

PUBLIC SEPTIC TANKS

6. Public Septic Tanks

Executive Summary

Public septic tanks were removed from the small sewage works econometric model in the Strategic

Review of Charges 2006-10 thereby removing any allowance for their operation. In the expectation that

the Commission will adopt the same approach in 2010-14, a special factor allowance is required to

address this. The special factor claim is valued at £1.2m (2007/08 prices), including power and general

and support expenditure; but excluding SEPA charges.


Background

Scottish Water operates 1,220 public septic tanks, approximately 62% of the total number of WWTW in

Scotland. Where the level of treatment is appropriate, septic tanks offer the most cost effective form of

treatment and so any change to the number of tanks by changing the treatment method would result in

an increase in overall cost.

Econometric Models

Public septic tanks were removed from the scope of the econometric models by the Commission in the

Strategic Review of Charges 2006-10 (SR06) Final Determination and replaced by an operating cost

allowance for their operation. It is expected that the Commission will adopt the same approach for the

Strategic Review of Charges 2010-14 (SR10) and the operating cost allowance will be recalculated.

Quantification

The current value of the claim is based on the figure in the 2008 Annual Return E table (E8) ‘Wastewater

Explanatory Factors – Sewage Treatment Works’. This has been reviewed by the Reporter. This gives the

direct costs for operation of all public septic tanks as £2.7m including general and support expenditure

(see table below). This figure also includes for SEPA service charges. SEPA charges are excluded from

the quantification of this special factor, as they are excluded from the econometric models. When SEPA

charges are removed, this brings the claim to £1.2m in 2007/08 prices.

Tankering costs are included in Table E10 of the Annual Return and modelled in the Sludge Econometric

model. Therefore these costs are not relevant to this Special Factor and are not included in the value of

the claim.

Scottish Water: Second Draft Business Plan. Public Septic Tanks

30

PUBLIC SEPTIC TANKS

Extract from Table E8 of Annual Return 2008

Septic Tanks 2007/08 Line Ref Unit

Number of septic tanks E8.8 1,220

Load (kg BOD/day) E8.18 7,191

Direct costs for all public septic tanks (£000s) E8.38 2,457

General and support expenditure (£000s) E8.39 251

Service charges SEPA (£000s) E8.42 1,467

Functional expenditure (£000s) E8.40 2,708

Claim: Functional expenditure less SEPA charges £1.2m


Opportunities for consolidation of numbers of tanks are rare, since assets are usually sited at the most

appropriate location for the community with respect to hydraulics and land availability. Any change to

the current system would most likely require the construction and operation of expensive pump transfer

systems, while there would be no reduction in the volume of sludge to be removed, as there would be no

change in the population served or the resultant sewage loading.

Across the Highlands, it is typical for septic tanks to be visited once per month where the operator will

take turbidity/SS samples and check the tanks for blockages to the inlet and outlet. There are some

tanks that are visited more frequently for specific reasons which include high population, tanks with

soak-away problems and tanks that discharge into sensitive waters. The average time spent servicing

septic tanks across the whole of Scotland has been calculated as 1.8 man days per tank per annum. This

figure is based on FTEs provided by Scottish Water’s Activity Based Management (ABM) system.

A comparison of figures from previous annual returns is given in the table below.

Comparison of Historical Septic Tank Costs

Costs (£,000) 2007/08 2006/07 2005/06 2004/05 2003/04 2002/03

Direct costs for all septic tanks 2,457 2,219 1,927 1,772 1,686 1,342

General and support expenditure 251 330 371 431 375 594

Functional expenditure 2,708 2,549 2,297 2,203 2,061 1,936

Service charges SEPA 1467 1,342 928 989 846 356

Functional expenditure excl SEPA (nominal prices)

1,241 1,207 1,369 1,214 1,215 1,580

Functional expenditure excl SEPA (2007/08 prices)

1,241 1,257 1,479 1,346 1,389 1,856

The table above shows how Scottish Water’s septic tank costs have trended in real terms since 2002/03

and have now established at £1.2m for functional expenditure excluding SEPA charges.

Conclusion

The value of this special factor claim excluding SEPA charges is £1.2m in 2007/08 prices.

Scottish Water: Second Draft Business Plan. Sewer Laterals

31

SEWER LATERALS

7. Sewer Laterals

Executive Summary

Due to a difference in sewerage legislation between Scotland and England and Wales (E&W), only Scottish

Water has the responsibility of operating and maintaining sewer laterals; a specific section of the sewer

network, in 2007/08. This special factor claim quantifies Scottish Water’s operational cost for its sewer

laterals. The claim is calculated as £7.2m in 2007/08 prices, reflecting the additional costs of

maintaining laterals.



A sewer lateral is a particular section of the sewerage network for which Scottish Water is responsible by

statute under the Sewerage (Scotland) Act (1968) and therefore Scottish Water’s responsibility for sewer

laterals is out with managerial control. By contrast, in E&W with certain exceptions it is the property

owners and not the water and sewerage companies that are presently responsible for maintaining their

sewer laterals. There is an ongoing consultation by DEFRA on the future transfer of responsibility for

sewer laterals in E&W to the water and sewerage companies. However, as the Commission benchmarks

Scottish Water against E&W in the base year (2007/08), this special factor is valid.

The Scottish definition of a sewer lateral is, per the Sewerage (Scotland) Act (1968), the section of a

sewer starting at the curtilage of a property and ending where it joins the public sewer. Under the Act

Scottish Water has statutory responsibility for their operation and maintenance.

In E&W, legislation distinguishes between lateral drains and private sewers. Together they form the

sewer lateral in the Scottish sense. Apart from the sewer laterals of detached houses, ownership of

sewer laterals was vested in the sewerage undertaker until 1936. Thus the sewer laterals built prior to

1937 remain the responsibility of the water and sewerage companies and are known as Section 24 sewers.

The Health Act, 1936, (now Water Industry Act 1991 and The Building Act 1984), made the property

owners responsible for their sewer laterals. The Water Act 2003 made the sewerage undertaker

responsible for providing new sewer laterals but enabled them to charge the property owner for provision

and maintenance costs.

Since privatisation, DEFRA has been consulting with the water industry on how and when the

responsibility for the sewer laterals should be transferred to the sewerage undertakers in England and

Wales.

Econometric Models

Since the water and sewerage companies in England and Wales are not responsible for the operation of

sewer laterals (with the exception of Section 24 sewers), the costs of sewer laterals are not fully

accounted for in the sewer networks estimate. Therefore, the relatively higher cost for their operation

is not forecast by this econometric model.

Scottish Water has re-calculated the Ofwat sewer network model including Scottish Water data (see

Annex 8). This model has the same functional form as Ofwat’s model and predicts operational costs


32

SEWER LATERALS

according to the length of sewer network. Including Scottish Water’s sewer lateral length within its

modelled sewer network length raises costs by £0.5m. This does not fully reflect the operational costs

incurred by Scottish Water with regard to sewer laterals.

Quantification

Introduction

Throughout the Strategic Review process, Scottish Water has considered a number of potential

approaches to calculating the value of our sewer lateral special factor claim. The approach we have

adopted is based on our own data, as this most accurately reflects the operating costs faced by Scottish

Water with respect to sewer laterals.

Operating cost data used

Scottish Water does not separately identify the costs of carrying out specific activities on sewer laterals.

We do, however, identify the total operating cost associated with all sewers. The quantification of the

cost of operating sewers is calculated from Activity Based Management (ABM) data. ABM provides

analysis of the costs of key activities and processes, and links these to the factors that cause or drive

Scottish Water’s level of operating cost. In this case, ABM has been used to summarise captured

operating costs for work relating to all sewers, including laterals (see the table below). The total

operating cost of all sewers was £19.3m in 2007/08.

Operating cost for all sewers from ABM, 2007/08

Activity Group Activity Total Cost (£000)

Opex repair of wastewater service connections 554

Opex repairs and maintenance of wastewater street furniture 195

Opex sewer repairs 7,141

Perform WW Infrastructure Investigations 237

Remove chokes 6,263

Sewer cleansing 1,331

Wastewater transportation

Sewer inspections including CCTV surveys 591

Handle Odour Complaints 450

Handle other Wastewater Complaints 714

Handle Sewer Flooding Complaints 1,570

Wastewater Complaint Investigation

Handle Wastewater Infrastructure Safety Complaints 226

Total 19,272

Cost drivers adopted

A proportion of the total operational cost of sewer networks can be attributed to operating sewer

laterals based on the use of cost drivers to apportion costs in a transparent manner. Two cost drivers

have been used:


33

SEWER LATERALS

• The ratio of the length of sewer mains to sewer laterals as reported in the 2007/08 Annual

Return.

• The proportion of call-outs attributed to sewers and sewer laterals, obtained from Scottish

Water’s Promise system.

Each of these cost drivers can provide an estimate of the costs attributable to sewer laterals. These

estimates are then reduced by 22%, the estimated proportion of sewer laterals operating cost attributed

to S24 sewers. Annex 9 details our approach to calculating this percentage. The cost estimates

associated with each cost driver are provided below.

Quantification - Length of Sewer as the Cost Driver

The length of sewers and laterals, as published in the 2007/08 Annual Return (lines E7.8 and E7.9), has

been used to split the sewer operating by main sewers and laterals (see table below).

Length of Sewer and Laterals in Scottish Water

Type Length (km) Proportion of total length of sewers

Estimate of cost* (£m)

Sewers only 33,451 67% 13.0

Lateral 16,312 33% 6.3

All sewers 49,763 100% 19.3

* Cost is split according to the proportion of sewer length of each type

The unadjusted Scottish Water estimate for operational costs associated with sewer laterals is £6.3m.

This value includes activity on sewer laterals that would be considered as Section 24 sewers under English

and Welsh legislation. Reducing this value by 22%, results in an adjusted value of £4.9m.

Scottish Water recognises that this approach assumes that the operational cost per kilometre of sewer

lateral or main sewer is the same. This may be a conservative assumption, as lateral chokes cause floods

more often than other sewers (see table below regarding call-outs). There are several reasons why this

may happen. The lateral is often a small diameter pipe laid at a low gradient with long periods of low

flow, sometimes with an interceptor close to the public sewer which increases the risk of blockage.

Invasion of tree roots can add to the increased frequency of lateral chokes. Some incidents with laterals

also require multiple callout to a single property, sometimes on a single day. Therefore laterals may be

relatively more expensive to repair than main sewers. Scottish Water does, however, note that as

laterals are often located away from highways, there is less need for traffic management and road

resurfacing during excavations and repair; therefore lower costs might be incurred.

Quantification - Callouts as the Cost Driver

The table below summarises 2007/08 information from Promise on sewer failure due to internal or

external flooding incidences. This illustrates the prevalence of sewer laterals failure relative to main

sewers. If the total cost of all sewer network activities was apportioned by the number of call-outs, the

unadjusted operating cost estimate for sewer laterals is £13.3m. Reducing this estimate by 22% to

account for Section 24 sewers gives an adjusted figure of £10.4m.


34

SEWER LATERALS

Call-outs requiring work on sewers and sewer laterals.

Type Number of call-outs requiring work on sewers and sewer laterals

Proportion of total number of

call-outs

Proportion of cost* (£m)

Sewers only 6,794 31% 6.0

Lateral 15,068 69% 13.3

All sewers 21,826 100% 19.3

* Cost is split according to the proportion of call-outs for each type of sewer

We note that this approach assumes that call-outs to laterals are as costly as call-outs to main sewers.

Overall value of the claim

The value of the claim submitted is based on the following:

• Section 24 adjusted sewer lateral operating costs using length as the cost driver - £4.9m

• Section 24 adjusted sewer lateral operating costs using call-outs as the cost driver - £10.4m

The average of these estimates is £7.7m. The final step is to reduce this by £0.5m, which reflects our

estimate of the sewer lateral operating cost attributed to Scottish Water based on the econometric

models (see Annex 8).

Consequently, the value of the sewer lateral claim is £7.2m.

Mitigating the impact

In line with best practice, Scottish Water minimises sewer lateral operational costs by optimising sewer

lateral activity.

• Call centre decision trees have been in place for several years and allow jobs to be

optimised:

o Most importantly, to ensure that the lateral is the responsibility of Scottish Water

and not a private drain for which the customer is responsible

o To determine whether jobs are urgent and need immediate attention (possibly

resulting in overtime hourly rates) or can be deferred until normal working hours

o To identify the type of problem so that a choke squad or a Network Service Operator

should be sent as appropriate

o To determine the number of operational staff required at a job (area and job

specific)

Previously, choke squads may have been sent out immediately but unnecessarily resulting in

increased costs and unavailability of squads to tackle urgent jobs.

• The Flood Management Action Plan (FMAP) Process has been in place since August 2008. All

internal flooding incidents are now investigated with a CCTV survey to determine the cause


35

SEWER LATERALS

(sewer incapacity, silted sewers, tree root obstruction). This determines where investment

is to be channelled to prevent future reoccurrences. The FMAP process is more streamlined

than its predecessor, the Internal Flooding Other Causes (IFOC) Programme which was

commenced in May 2006. This also investigated the causes of internal flooding to determine

where investment should be made.

Conclusion

Scottish Water has the additional responsibility of operating and maintaining sewer laterals due to

legislative differences between Scotland and England and Wales. Operating expenditure for these

laterals above that taken account of in the econometric models is £7.2m in 2007/08 prices.


36

SEWER LATERALS

Annex 8: Adjustment to Claim to Reflect Operational Costs Predicted by

Sewer Networks Model for Lateral Sewers

Scottish Water’s sewer lateral length is included within its modelled sewer network length. Therefore

the claim is reduced by the value of operational costs the sewer networks model predicts for non section

24 sewer laterals (i.e. sewer laterals built from 1937 that would not be the responsibility of Scottish

Water if legislation from England and Wales applied).

The sewer networks model predicts it should cost Scottish Water:

• £34.2m to operate all of its sewers (33,514km main sewers and 16,312km of sewer laterals)

• £33.7m to operate main sewers and Section 24 sewers (33,514km plus 3,534km of S24 sewer

laterals)

Therefore the claim should be reduced by the £0.5m difference in these two values.

These values are determined by Scottish Water’s re-calculation of Ofwat’s 2007/08 sewer networks

model, published at end January 2009. This model includes Scottish Water data when estimating the

model coefficients.


37

SEWER LATERALS

Annex 9: Sewer Lateral Lengths Used in the Special Factor Claim

Adjustment for Section 24 Sewers

Based on Annual Return data, the total length of sewer laterals under Scottish Water’s responsibility is

16,312km. However, this includes for sewers constructed prior to 1937 known as Section 24 (S24) sewers

in England and Wales, which are already the responsibility of the WaSCs in England and Wales.

Pre 1937 sewers must be accounted for in a special factor submitted, i.e.:

Sewer lateral length used for the special factor claim

=

Scottish Water’s sewer lateral length less assumed Scottish S24 length

The methodology uses property data to determine the percentage of non-S24 sewers in Scotland. This

percentage is then used to reduce the operating cost estimates for sewer laterals to ensure these costs

are not included in the claim.

In the absence of better data the proportion of properties built prior to 1937 has been used to

approximate for the proportion of sewer laterals that would be considered as Section 24 sewers in

England and Wales. The approach adopted is outlined below.

Step 1 - Scottish House Condition Survey (SHCS) data (2005/06 findings) used for this claim is presented

in age band in the table below.

Number of properties in Scotland

Age Band House type

Number of houses in Scotland

(000s)

Other Flats 41

Tenement 152

Detached 90

Semi-detached 60

Pre 1850 to 1919

Terraced 46

Total pre 1850 to 1919 389

Other Flats 98

Tenement 40

Detached 33

Semi-detached 83

1919-1944

Terraced 43

Total 1919 to 1944 297

1945 onwards All types 1,629

Total (up to present day) All 2,315

Step 2 - The SHCS team, from the Scottish Government, provided Scottish Water with further data in

order to better determine the percentage of houses built between 1919 and 1944. This data is presented

below.


38

SEWER LATERALS

Additional Data from SHCS team

Revised SHCS Data

Age Category Number of Buildings Built

(000s)

pre 1919 385

1919-1929 66

1930-1949 326

post 1949 1,537

Total 2,315

Step 3 - Using the additional information provided by the SHCS team and the 1919 to 1944 total number

of properties in Scotland of 297,00 (from the previous table), the estimated number of properties built

between 1945 and 1949 is 95,000 (326,000 + 66,000 - 297,000). This leaves 231,000 properties built

between 1930 and 1944. In order to derive this figure, we have assumed that properties were built at a

constant annual rate through the 1930s and that there was no house building during the war years (1940 –

1945)8. This equates to 23,100 properties per year built during the 1930s.

Step 4 - The additional information provided by the SHCS team, along with the assumption that 23,100

properties were built per year in the 1930 to 1939 period are used in the table below to estimate the

percentage of houses built between 1937-44 in the period 1919–1944.

Calculation of percentage of houses built between 1937 and 1944 in the period 1919–1944

Housing stock built in period 1919 to 1944 (000s) [1]

Housing stock built between 1937 and 1944 (000s) [2]

Percentage of housing stock 1919-1944 built between 1937 and

1944[3] = [2] / [1]

66 + 231 = 297 23.1 x 3 years = 69.3 23.3%

Step 5 - The initial data provided by SCHS is reduced by the number of properties that would be served

by the Scottish equivalent of S24 sewers (i.e. sewer laterals constructed prior to 1937). This gives the

number of properties not connected to main sewers by S24 sewers.

For pre 1850 to 1919 all house types, other than detached9, are affected by S24. For the period 1919 to

1944, 23.3% of all house types, other than detached, are assumed not to be affected by S24. For the

remaining age bands, no properties are affected by S24. This calculation is shown in the table below.

8 The impact on house building from the war was assumed to start in 1940 rather than in 1939. 9 The sewer laterals of detached houses have never been the responsibility of WaSCs in England and Wales and therefore the

operation and maintenance of sewer laterals of detached houses in all age bands represent a difference in responsibility between Scotland and E&W. These are not included in the estimate of Scottish Water’s S24 sewers length.


39

SEWER LATERALS

Number of properties in Scotland (Pre 1850 – 1944)

Age Band House type Number of houses in

Scotland Number of houses not connected by S24 sewer equivalent (000s)

Other Flats 41 0

Tenement 152 0

Detached 90 90

Semi-detached 60 0

Pre 1850 to 1919

Terraced 46 0

Total pre 1850 to 1919 389 90

Other Flats 98 23

Tenement 40 9

Detached 33 33

Semi-detached 83 19

1919-1944

Terraced 43 10

Total 1919-1944 297 94

1945 onwards All types 1,629 1,629

Total (up to present day) All 2,315 1,863 (78%)

78% of properties in Scotland are connected via a sewer lateral operated by Scottish Water that would

otherwise be the responsibility of the home owner under the S24 legislation in England and Wales i.e.

detached or were built in 1937 or subsequently. Therefore approximately 78% of Scottish Water’s sewer

lateral length should be reflected within the special factor claim. Consequently, 78% of the sewer lateral

operating cost estimates provided in the main text are used when deriving the final claim value10.

10 Variations relate to rounding and approximations made in the calculations.

Scottish Water: Second Draft Business Plan. Supply of Chemicals

40

SUPPLY OF CHEMICALS

8. Supply of Chemicals

Executive Summary

Due to Scotland’s sparse nature, remote rural areas and distances from centres of manufacture, Scottish

Water’s suppliers charge Scottish Water higher transportation costs for chemicals. Furthermore, as a

result of the high number of small water treatment works (WTW), chemicals cannot be supplied in bulk

therefore the cost per tonne is greater.

This special factor claim quantifies the additional costs to Scottish Water not faced by water companies

in England and Wales (E&W) and not reflected in the econometric models. The claim is for £0.3m in

2007/08 prices.



Scottish Water provides water and sewerage services across the most sparsely populated areas of Britain.

Water and wastewater treatment works are situated at appropriate locations in order to serve the

population. Chemicals are required in the treatment process of Scottish Water’s treatment works. As

well as the sparse distribution of assets, the size of Scotland means that the majority of chemicals are

required to be delivered to locations at great distances from centres of manufacture. Scottish Water’s

paper ‘Population Sparsity – Determining a Proxy for E&W within Scotland’ demonstrates that its

population is more sparse than that of the English and Welsh companies.

Scottish Water operates 313 WTW of which 261 WTW treat up to 10Ml/d. An average of 15,906 people

are served per WTW. Only Wessex Water has a smaller average number of people served per WTW (see

‘Population Sparsity – Determining a Proxy for E&W within Scotland’). As a result of the high number of

small water treatment works, chemicals cannot be supplied in bulk therefore the cost per tonne is

greater.

Size band of WTW Number of WTW

<=1 Ml/d 186

>1 - <=2.5 Ml/d 25

>2.5 - <=5 Ml/d 30

>5 - <=10 Ml/d 20

>10 - <=25 Ml/d 23

>25 - <=50 Ml/d 13

>50 - <=100 Ml/d 9

>100 - <=175 Ml/d 4

>175 Ml/d 3

Total number of works 313

This supply of chemicals special factor claim is for the excess cost of supplying chemicals over those

incurred by water companies in E&W. The extra costs are outside management control as Scottish Water


41

SUPPLY OF CHEMICALS

cannot influence the transportation charges imposed by its suppliers and in the provision of a Scotland-

wide service it cannot exclude costly, remote customers from its customer base.

Econometric Models

The claim reflects the additional costs of supplying two key chemicals, Aluminium Sulphate (Alum) and

Calcium Hydroxide (Lime), to water treatment works (WTW). Therefore the relevant costs appear in the

water resources and treatment model:

R&T functional expenditure = constant + a*(number of sources) – b*propn supply from boreholes

Resident winter population Distribution input

Modelled resources and treatment functional expenditure excludes power and SEPA costs

The variable of the number of sources divided by total distribution input reflects the fact that a company

that supplies its customers from many small sources is predicted to require additional opex. It should be

noted that this ratio is independent of the area served. Two companies could be identical in terms of

distribution input and number of sources but differ substantially in terms of area but this difference

would not be directly reflected in the model and therefore the predicted costs.

Annex 10 demonstrates that Scottish Water has a ratio of number of sources to distribution input that is

close to the E&W average. Because Scottish Water operates in an extremely rural environment and has

to deliver water to very dispersed populations, one would tend to expect a much higher ratio. What this

demonstrates is that Scottish Water has been able to achieve above-average efficiencies through the

consolidation of sources, given that it achieves similar values of sources per DI as far less rural companies

in England.

The crucial difference between E&W companies and Scottish Water is that sources of the same average

size have to be supplied in a far larger area in Scotland. As noted above, the economies-of-scale variable

does not capture these additional costs of remoteness because it does not directly account for area.

Hence the model does not reflect variations in supply costs of chemicals caused by sparsely located

WTWs where transport costs for delivery are higher.

Quantification

The claim for this special factor only includes alum and lime. The annual cost of the supply of these

chemicals for the financial year 2007/08 has been provided. The cost of annual supply for Alum and Lime

are:

• Aluminium Sulphate (Alum): £3.8m, Calcium Hydroxide (Lime): £1.1m

Details of the costs in supplying chemicals to WTW in E&W are not available in sufficient detail to allow

comparison with Scottish Water’s own costs. For this reason a comparison has been made within Scottish

Water between a Scottish area (the proxy area) deemed to be suitably equivalent to E&W, and the

remaining, more rural areas. The most appropriate combination of regions to be considered as the proxy

area for E&W is Ayr, Clyde and Tweed. The analysis behind the establishment of the proxy area is

presented in the ‘Population Sparsity – Determining a Proxy for E&W within Scotland’ report.


42

SUPPLY OF CHEMICALS

Scottish Water data on the cost of supplying chemicals per tonne is available by WTW and hence by

region. Therefore a comparison can be undertaken between the typical cost of delivering chemicals per

tonne within the Scottish Water proxy region and the other more rural regions.

Claim for Additional Cost of Supplying Chemicals to Remote Scottish Water Regions

ALUM LIME

Average Rate

(£/tonne)

Amount (1000 tonnes)

Cost (£000s)

Cost @ Proxy Rate (£000s)

Average Rate

(£/tonne)

Amount (1000 tonnes)

Cost (£000s)

Cost @ Proxy Rate (£000s)

Ayr 88 10.8 943 943 92 2.2 205 203

Clyde 88 7.4 648 649 89 2.1 186 189

Tweed 88 1.8 161 161 92 1.0 94 93

Ness 129* 1.6* 212 144 142 0.4 63 40

Forth 91* 7.0* 637 612 95 1.1 101 96

Tay 97* 4.0* 389 353 94 1.3 118 114

Don 112* 3.3* 367 286 112 1.3 145 117

Nith 87 4.9 422 425 94 1.1 101 98

Proxy 88 91

Total (£000s) 3,781 3,575 1,013 951

Variance (£000s)

206 62

Total Claim (£000s)

268

* The true volumes and costs have been adjusted downwards in these regions to reflect that certain WTW

are supplied with alum at greater concentrations.

The costs per tonne vary across the proxy regions and so a weighted average is used based upon volume.

The average cost per tonne to supply the chemicals in the proxy regions is then applied to the remaining

regions to calculate the cost at the proxy rate. The claim equates to a cost that would be saved if the

two chemicals listed were supplied to the more rural regions at the same delivery rates as the Scottish

Water proxy region for E&W.

It is apparent from the table above that it is cheaper to supply chemicals to the proxy area than the rural

regions in Scotland. For alum, the average rate of supply in the proxy area is £88 per tonne which is

compared with £129 and £112 per tonne in Ness and Don respectively, which are the two most rural

regions in Scotland.

Mitigating the impact

Scottish Water undertakes a tendering procedure for supply of chemicals to ensure that the lowest cost is

achieved.

Conclusion

Due to Scotland’s sparse nature, remote rural areas and distances from centres of manufacture, Scottish

Water incurs greater costs in the transportation and supply of chemicals. This special factor claim

quantifies the additional costs to Scottish Water not faced by water companies in E&W and not reflected

in the econometric models. The claim is for £0.3m in 2007/08 prices.


43

SUPPLY OF CHEMICALS

Annex 10: Comparison of Resources & Treatment model:

Number sources / distribution input

Number of sources

Distribution input (Ml/d)

Number of sources / D.I.

Folkestone & Dover 19 44 0.43

Mid Kent 57 155 0.37

Cambridge 25 73 0.34

Wessex 100 353 0.28

South East 106 377 0.28

Tendring Hundred 8 29 0.27

Sutton & E Surrey 35 151 0.23

Southern 119 562 0.21

Anglian 226 1,165 0.19

Dee Valley 12 66 0.18

Scottish Water 371 2,271 0.16

Yorkshire 163 1,280 0.13

South West 54 435 0.12

Portsmouth 22 178 0.12

Three Valleys 102 835 0.12

Dwr Cymru 96 840 0.11

Severn Trent 174 1,883 0.09

Northumbrian 103 1,155 0.09

South Staffs 26 318 0.08

United Utilities 151 1,849 0.08

B&W Hants 10 153 0.07

Bristol 18 282 0.06

Thames 136 2,572 0.05

The above table demonstrates that Scottish Water is in the middle of the pack in respect of the variable

in the resources and treatment model: number of sources / distribution input. This is not what would be

expected from the most rural water company in Britain. The above demonstrates that there is no

particular allowance made in the model for the rurality of Scottish Water’s operational area. The claim

is not for any extra dis-economies of scale costs but for the extra costs of supplying works (be they large

or small) in remote areas.

Scottish Water: Second Draft Business Plan. Travel

44

TRAVEL

9. Travel

Executive Summary

Due to the scale and geography of Scotland, its sparse population and, consequently, the additional

number of assets required to provide water and waste water services, Scottish Water incurs additional

travel costs in operating its assets. These additional costs are outside management control, not faced by

water companies in England and Wales (E&W) and are not reflected in the econometric models. The

claim is for £9.1m in 2007/08 prices, as broken down in the table below.

Summary of additional travel costs incurred by Scottish Water (2007/08 prices)

Claim Operational

Travel (£m p.a.)

Labour time lost due to additional travel 6.1

Excess fuel and mileage costs 2.2

Excess cost of vehicle repairs and maintenance 0.8

Total 9.1



Scottish Water provides water and sewerage services across the most sparsely populated areas of Britain.

It serves the largest area of all water companies, almost three times the size of the next largest

company, Anglian, and five times larger than the average. As a result, Scottish Water employees must

travel considerable distances between the assets required to serve the sparsely distributed population

which leads to a number of additional costs, comprising:

• The additional time employees spend travelling between assets.

• The additional fuel and maintenance costs associated with travelling long distances.

This travel special factor claim is for excess travel costs incurred by Scottish Water for operational travel

over those incurred by water companies in E&W. The extra costs of sparsity are a result of Scottish

Water’s operational area and are outside management control in so far as Scottish Water has to provide

universal access and cannot exclude costly, remote customers from its customer base.

Econometric Models

The econometric models do not explicitly include travel cost drivers. Instead, travel costs are a

component of the rolled up operational costs included in the models. Some of the models could be

interpreted as reflecting population spread but their relation to company-wide travel costs are rather

ambiguous.


45

TRAVEL

Do the Efficiency Models Reflect Variation in Travel Costs?

Model Travel costs

included in model Variables

Does the Variable Explain Travel in rural areas

Water distribution Travel to repair leaks and other network activities

% mains > 300mm

No, as model reflects additional costs in urban areas – large trunk mains tend to occur in more urban areas, and companies with a high proportion of large trunk mains are predicted higher opex. It does not predict additional opex for more rural companies who have to drive greater distances to maintain water mains.

Water resources and treatment

Travel to raw water sources and treatment works

Number sources/D.I. & proportion D.I. from boreholes

No, as the variables are not related to travel. See Annex 10 Supply of Chemicals report for comparison of number sources/D.I.

Water power None n/a n/a

Water business activities

Sampling, non-ops travel

Number of billed properties

No. Companies could have similar property numbers but very different pop sparsity & hence asset mix.

Sewerage network

Travel to carry out maintenance and repairs to the sewer network

Holiday:resident pop, pop/km mains,

km mains/area served

Whilst the area served per km of sewer could be expected to act as a proxy for dispersal of assets, and so travel requirements, the population per km is clearly a measure of population density. The coefficient on population per km is larger than the other co-efficients in the model so, as this is a log model, has a greater impact suggesting the model penalises companies with widely dispersed populations.

Large sewage treatment works

Travel to treatment works

Load & tight BOD / suspended solids consents

No, as the model is based only on the volume of treatment and does not take into account variations in pop density.

Small sewage treatment works unit cost

Travel to treatment works

Load and level of treatment

No, as the higher unit costs the model estimates for smaller WWTW are reflective of E&W costs and so do not reflect the extent of travel in Scotland.

Sludge treatment and disposal unit cost

Sludge transportation

Volume of sludge, disposal route

No, as the unit cost estimated by the model is based upon the E&W average from which SW is different. No scale of sludge centre or distance in sludge transportation is factored in.

Sewerage business activities unit cost

Sampling, non-ops travel

Number of billed properties

No. Companies could have similar property numbers but very different pop sparsity & hence asset mix.

The ‘Population Sparsity – Determining a Proxy for E&W within Scotland’ report demonstrates that the

population in Scotland is more sparse than in the areas served by the English and Welsh companies. The

econometric models are biased towards the average E&W aspect in terms of costs and activities

undertaken. Thus, they do not adequately reflect the additional Scottish Water travel costs incurred as a

result of the differing nature of the spread of population.

In 2003/04, Ofwat accepted a special factor submitted by Thames Water for the excess travel costs as a

result of congestion. Amongst E&W companies, Thames Water is an outlier due to its significantly higher

population density. Ofwat’s acceptance of Thames Water’s claim demonstrates that the regulator

recognises that extra density-related costs can occur, which are not captured in the econometric models.


46

TRAVEL

Just as Thames Water’s costs do not conform to the functional form of the econometric estimates, so do

Scottish Water’s costs exceed the predicted values.

Quantification

The travel special factor claim covers the additional costs associated with:

• Operational staff time lost due to travelling between Scottish Water assets: £6.1m.

• Fuel costs: £2.2m.

• Vehicle repair and maintenance costs: £0.8m.

Scottish Water Proxy for England and Wales

The necessary level of detailed travel data is not available for the water companies in E&W to allow the

claim to be quantified in relation to the costs incurred in E&W. For this reason a proxy for E&W has been

developed, using a combination of several of Scottish Water’s regions, that can be considered to be

reflective of the average situation in E&W, across a number of relevant factors. The analysis behind the

establishment of the proxy area is presented in the ‘Population Sparsity – Determining a Proxy for E&W

within Scotland’ report.

The proxy area, which is a combination of the operating areas of Ayr, Clyde and Tweed, has a similar

urban:rural balance to the most rural companies in E&W. As a result, this claim is on the conservative

side since it is quantified against a base measure which is more-rural than the average experienced in

E&W.

The claim is quantified as the difference between the existing, relevant, Scottish Water costs and the

costs that would be incurred by Scottish Water if the costs in the proxy area applied across all of

Scotland. The three factors (additional travel time, additional fuel costs and additional vehicle repair

and maintenance costs) are established on a per-property basis for the proxy area. These property-based

factors for the proxy area are then applied to the total of properties in Scotland, to establish the costs

that would have been incurred if the E&W proxy area was representative of all of Scottish Water.

The number of properties is considered an appropriate measure of scale as the location of Scottish

Water’s customers is outside managerial control and, furthermore, the number of properties is (or is

correlated to) a cost driver in almost all of the econometric models. The extent to which travel costs

vary on a per property basis broadly indicates the extent to which they are not accounted for in the

models. The table in Annex 12 considers the appropriateness of using properties to scale against each of

the models which contain travel costs.

Methodology: Claim for Labour Time Lost

Scottish Water’s operational staff spend a significant portion of their working day travelling between

assets, particularly in the Highlands and Islands. Generally using Scottish Water vehicles, their travelling

time is captured by telematics, a vehicle-tracking software system (details in Annex 11). The travel time

of operational staff using their own vehicles has been calculated from mileage expense claims using an


47

TRAVEL

average speed of 42 mph (refer to Annex 13 for average speed data); this approximation contributes just

6% of operational travel time.

This travel time information has been provided for each of Scottish Water’s operating regions so that a

comparison can be undertaken between typical travel time of operational staff within the Scottish Water

proxy region and the other more rural regions. As the regions are of different sizes, the travel data has

been scaled using the number of properties to allow comparison, i.e. travel time per property served in

each region.

The table below presents the calculation for excess operational labour time lost through travel compared

with E&W, using the Ayr, Clyde and Tweed as a proxy for E&W. The claim equates to the cost of the

labour time that would be saved if the more rural regions were able to operate with the same travel time

per property served as the Scottish Water proxy region for E&W.

The average travel time (per property) within the proxy area, is 0.21 hours per property (line E – 239

thousand hours divided by 1,130 thousand properties). This value is an estimate of the average travel

time per property served in E&W. Applying the 0.21 factor to the total properties (2,394,000) gives rise

to a normalised travel time of 507,000 hours across Scotland (line F). When deducted from the total

number of hours spent travelling (approximately 727,000 hours – line C), this represents 220,000 excess

hours spent travelling (line G), over the proxy area (and therefore E&W). Based on an average hourly

rate of £27.89, (see Annex 14 for details), the claim for the excess cost of operational travel time over

E&W amounts to £6.14m.

Value of Excess Labour Time Lost due to Operational Travel

Ref Don Forth Ness Nith Tay Proxy Area Total*

Total

A

Actual time spent travelling in fleet vehicles (000 hrs) 91 81 106 89 92 227 685

B

Estimated time spent travelling for operational staff using own vehicles (000 hrs)

5 5 8 6 6 12 42

C Total time spent travelling (000 hrs) [A+B]

96 85 113 95 98 239 727

D Connected Properties (000s) 245 215 99 350 356 1,130 2,394

E Travel time per connected property (hrs/per prop) [C/D]

0.39 0.40 1.15 0.27 0.28 0.21 0.30

F

Derived time spent travelling (000 hrs) (applying proxy time spent travelling per connected property) [D*EPROXY]

51.8 45.5 20.9 74.2 75.4 239.3 507

G Difference between actual and derived travel time (000 hrs) [C-F]

44 40 93 20 23 0 220

H (Average) of internal labour recharge rates (£/hr) 27.89

I Value of labour time lost due to excess operational travel (£m) [GTOTAL*H] 6.14

* The proxy area consists of Ayr, Clyde and Tweed


48

TRAVEL

The table above demonstrates that the most travel is undertaken in the Ness region. The notional time

spent travelling in the Ness region is 1.15 hours per connected property which is almost three times more

than that of the region with the next highest time spent travelling per property served. The excess

travel in the Ness region is the largest contributor to the value of the claim. This is in line with the

findings in the ‘Population Sparsity – Determining a Proxy for E&W within Scotland’ report which outlines

that as well as being the largest area, the population and works are sparsely distributed.

Methodology: Excess Fuel Costs and Vehicle Expenses

Fuel costs and vehicle expense claims are a function of distance travelled (mileage).

• Fuel costs are derived from fuel cards used by Scottish Water liveried vehicles to pay for

fuel.

• Vehicle expenses are claimed by employees for journeys made with their own vehicles. The

current mileage rate used by Scottish Water is 40p per mile11. The mileage rate covers the

costs of running and maintaining the vehicle, such as depreciation, fuel, oil, servicing,

repairs, insurance, vehicle excise duty and MOT.

Fuel costs and vehicle expenses, assigned to operational regions, are obtained from Scottish Water’s

financial systems. The quantification of the claim follows a similar methodology used in previous

sections where fuel costs and mileage expenses per property served for each region are compared with

that of the Scottish Water proxy area for E&W. The table below presents the calculation for excess fuel

and mileage costs compared with E&W, using the Ayr, Clyde and Tweed as a proxy for E&W.

The average fuel and mileage cost in the proxy area is £1.23 per property (line E). This value is an

estimate of the fuel and mileage cost per property served by the companies in E&W. Applying the £1.23

per property to the total number of properties (2,394,000) gives rise to a normalised cost of £2.917m

(line F), which is £2.214m lower than the current pan-Scotland figure of £5.131m. Therefore the excess

fuel and mileage costs over E&W is estimated at £2.2m.

11 This rate is in line with HM Revenue and Customs statutory mileage rates from 2002/03 onwards.


49

TRAVEL

Claim for Excess Fuel and Mileage Costs for Operational Travel

Ref Don Forth Ness Nith Tay Proxy Area Total*

Total

A Fuel cost (£K) 628 584 840 592 657 1,159 4,459

B Mileage costs (£K) 77 62 132 111 72 218 672

C Total fuel and mileage costs (£K) [A+B]

705 646 972 702 729 1,377 5,131

D Connected Properties (000) 245 214 99 350 356 1,130 2,394

E Cost per connected property (£) [C/D]

2.88 3.01 9.86 2.01 2.05 1.22 2.14

F Expected Cost (applying Proxy unit cost) (£K) [D*EPROXY]

298 262 120 427 434 1,377 2,917

G Difference between actual and expected amount (£K) [C-F]

407 384 852 276 295 0 2,214

H Value of Ops Travel claim for fuel and mileage costs (£M) 2.21

* The proxy area consists of Ayr, Clyde and Tweed

Again, the table above demonstrates that the most fuel and mileage costs are incurred in the Ness

region. The average cost in the Ness region is £10 per connected property which is over three times

more than that of the region with the second highest cost per property served. The excess fuel and

mileage cost in the Ness region is the largest contributor to the value of the claim.

Methodology: Excess Vehicle Repairs and Maintenance Costs

Over 2007/08, Scottish Water’s vehicle maintenance and repair costs amounted to £2.7m12. These costs

are a function of distance travelled as vehicles are serviced based on mileage. Additional vehicle repair

and maintenance costs are incurred by Scottish Water due to the necessity to operate and maintain

assets in rural areas.

The following is the calculation of Scottish Water’s excess vehicle maintenance and repair costs over

E&W, using the Ayr, Clyde and Tweed as a proxy for E&W. The claim equates to the estimated cost of

vehicle maintenance and repair that would be saved if the number of miles covered per property in the

more rural regions was the same as that of the Scottish Water proxy region for E&W.

The average mileage per property is much higher in Scottish Water’s rural areas than in the proxy area as

shown in the table below. The number of miles in the non-proxy area (1,264) is multiplied by this

difference of 4.8miles per property to give 6million excess miles in Scotland over E&W.

12 Taken from Scottish Water Fleet invoices


50

TRAVEL

Mileage per Property

Mileage (000s)

No. of properties (000s)

Mileage per property

Scotland 18,124 2,394 7.6

Proxy Area 5,685 1,130 5.0

Non-proxy Area 12,440 1,264 9.8

Difference between proxy area and non-proxy area

6,755 134 4.8

Excess mileage in Scotland over E&W (based upon proxy area)

1,263,800 properties x 4.8miles per prop

= 6,081,000 miles

SW maintenance cost per mile

£2.7m/18m miles = £0.13/mile

Costs of excess vehicle maintenance & repair

£0.13 per mile x 6m miles = £0.8m

The most mileage is carried out in the sparsely populated Ness region. Therefore, the Ness region is

again the highest contributor to the claim.


Scottish Water employs a number of methods to mitigate the impact of the amount of travelling

necessary in rural areas.

Local Training

It is the general policy of Scottish Water to send trainers to outlying/remote offices rather than staff

travel to central locations. As a general rule no staff should have to travel more than one hour to get to

a training centre. Exceptions are that staff based in Dumfries and Galloway must travel to Glasgow and

staff based in Campbeltown must travel to Oban.

Where training takes place in the Highlands and Islands, trainers are scheduled to undertake a number of

courses over several days to reduce travel costs. In 2007/08, 671 training courses have been held outside

the central belt (this includes Aberdeen and Inverness offices).

Scottish Water is now outsourcing driver training. This will be rolled out by the end of 2008 and though

concentrating on Health and Safety will consider efficiency. An online course will first identify high risk

drivers who will then undergo further training.

Some events requiring specialist facilities must be run centrally

Cross-Functional Working

Cross-functional working has been implemented in the more remote areas within Scotland. The ability

for staff to operate flexibly across operational boundaries helps reduce mileage travelled to treatment

works or parts of the network requiring maintenance. A multi-functional operator who is ‘on call’ can


51

TRAVEL

attend to any type of incident in a particular area, avoiding the need for a specialist operator to travel a

greater distance to attend.

Manpower Efficient Assets

In rural areas, it is often more cost effective to install treatment processes which require minimum

manpower intervention e.g. membrane filtration and septic tanks. Scottish Water has a number of

membrane plants throughout the rural areas of Scotland. 32 of 126 WTWs in the Ness Region are

membrane plants with 8 more planned for completion in the 2006 – 2010 period. Membrane plants only

require to be visited twice per week compared with other types of treatment works which are required

to be visited daily or manned full time.

On the wastewater side, where the level of treatment is appropriate, septic tanks offer the most cost

effective form of treatment. Septic tanks require minimal operational input and are typically emptied

once every six months. Scottish Water operates 1,220 septic tanks, predominantly in the rural areas of

Scotland.

Other General Activities to Minimise Travel

Scottish Water has implemented communication procedures to ensure that the costs of travel are

minimised across the business. These include:

• Participation in teleconferencing and video-conferencing to carry out meetings and training,

where feasible. Video facilities are available in 22 offices including all main offices, and

outlying offices including Lerwick, Kirkwall, Portree, Argyll, Stornoway, Caithness, and

Galashiels.

• Avoidance of non-essential meetings.

• Organisation of events locally rather than centrally.

• Encouraging car sharing and public transport

Fuel and Maintenance Cost Reduction Initiatives

Scottish Water has a number of initiatives to help minimise fuel and maintenance costs including

• Highlighting low cost fuel suppliers

• Issuing fuel saving tips

• Driver monitoring using telematics to identify sub-optimal use of vehicles.

• The commercial vehicle management and maintenance contract was re-tendered in July

2007.


52

TRAVEL

Excluded Activities

To avoid the possibility of double counting it is important that there is no overlap with other special

factors when quantifying the cost of the claim for the travel special factor. For this reason, all vehicles

that are used for cryptosporidium sampling were identified and removed as the travel cost for this will be

included in the crypto special factor.

Conclusion

Scottish Water incurs additional travel costs which are not faced by water companies in E&W as a result

of Scotland’s sparse population and large size. These additional costs are not reflected in the

econometric models. A special factor claim in the value of £8.0m in 2007/08 prices is submitted to allow

for this.

Summary of additional travel costs incurred by Scottish Water (2007/08 prices)

Claim Operational

Travel (£m p.a.)

Labour time lost due to travel 6.1

Excess fuel and mileage costs 2.2

Excess cost of vehicle repairs and maintenance 0.8

Total 9.1


53

TRAVEL

Annex 11: Telematics

The telematics data used in the analysis covers the 2007/08 financial year. Telematics is a vehicle

tracking software system which monitors approximately 95% of Scottish Water’s fleet of liveried vehicles.

Telematics is capable of monitoring the vehicle user, location, mileage, utilisation and fuel consumption.

Telematics records both the time when the vehicle ignition is switched on and any excessive idling when

a vehicle is stationary for longer than 10 minutes. The data used in sections of this report is the driving

time (time spent travelling) which is calculated by subtracting the excessive idling time from the time

when the vehicle ignition is switched on.

The data is extracted and allocated to regions directly from telematics historical data. Where a vehicle

was not assigned to a region it was allocated to the region where the majority of travel took place or

where the vehicle is based. A large number of vehicles are based at the Scottish Water depots at

Balmore Road, Glasgow (Clyde region) and Muir Road, Livingston (Tweed region). Although the

telematics data is recorded against these regions, some of these vehicles operate elsewhere. We are

unable to identify all the vehicles affected by this issue and so have made an assumption that they all

operate within the region that they are based. This has the effect of undervaluing the claim by

allocating more time to the proxy area (since both Clyde and Tweed regions are part of the proxy area).

Therefore, although the recorded travel time and mileage may be higher than the actual time and

mileage undertaken in the regions, the Scottish Water claim is underestimated as a result.

Operational mileage taken from telematics is also used when calculating excess maintenance costs.


54

TRAVEL

Annex 12: Scaling Factors Considered for the Travel Special Factor

As the regions are of different sizes and serve different numbers of properties, the travel data has been

scaled. It is important that the scaling factor is closely related (i.e. correlated) with the various

measures of scale used in the models. Each of the models (with the exception of the power model) will

account for an element of travel costs relative to the measure of scale used in that model. For the

purposes of the travel special factor, it is necessary to identify the element of travel not accounted for

by the measure of scale; i.e. there is only a coarse relationship between travel time and population

served or the number of properties served. However, a model using these variables as scaling factors

does not discriminate between variations in population density, and so does not accommodate all

variability in travel time.

Therefore the travel time data has been scaled by dividing by the number of properties to allow

comparison with the proxy area, e.g. travel time per property served in each region. The extent to

which travel costs vary on a per property basis broadly indicates the extent to which they are not

accounted for in the models.

The following table shows that the number of properties is highly correlated to the measures of scale

used in most of the models (or is in fact, the measure of scale) and is therefore an appropriate scaling

variable.

Options for Scaling between Regions to Quantify Travel Claim

Model Measure of Scale used in Model Relation to Number of Properties

Water distribution Winter population Highly positively correlated with the number of properties

Water resources and treatment Winter population Highly positively correlated with the number of properties

Water power NA Not relevant as there are no travel costs associated with power expenditure

Water business activities Number of billed properties Exact match

Sewerage network Kilometre of sewer Low correlation, however this is the only model for which this is the case

Large sewage treatment works Load Function of population equivalent and hence positively correlated with property numbers

Small sewage treatment works unit cost

Load (kg BOD5/day) Function of population equivalent and hence positively correlated with property numbers

Sludge treatment and disposal unit cost

Load (thousand tonnes of dry solids)

Function of population equivalent and hence positively correlated with property numbers

Sewerage business activities unit cost

Number of billed properties Exact match

The table below considers alternative scaling factors and demonstrates that the number of properties or

population is suitable measures of scale. The number of properties is considered an appropriate scaling

factor as the location of Scottish Water’s customers is outside managerial control and is (or is correlated

to) a cost driver in almost all of the econometric models.


55

TRAVEL

Evaluation of the Alternative Scaling Factors

Scaling factor Pros Cons

No. of properties • Location of SW customers is out with

managerial control. • No. of properties is (or is positively

correlated to) a driver in almost all of the econometric models.

• A property and hence the service it requires from Scottish Water is the same (more or less) regardless of location.

• Includes business customers. • Robust data available from Annual

Returns.

• Not as closely related to some of the models’ measures of scale as population.

Population • Similar impact on claim as number of

properties. • Closely related to those models that use

volume to scale. • Robust data available from Annual

Returns.

• Ignores business customers. • Possibly the number of connections is a

greater driver of SW’s cost rather than population. There could be variances in occupancy rates across the country.

Number of WTW

& WWTWs

• A significant portion of travel is in relation to visiting treatment works.

• Robust data available from Annual Returns.

• Treatment works vary in size, population served and the level of necessary manpower intervention therefore are not easily comparable. This could cause complication in the calculation.

• There is a significant volume of travel related to infrastructure assets.

• Not a measure of scale in the econometric models (with the exception of the large wastewater treatment works model).

Area • Robust data available from Annual

Returns. • Area is positively correlated with distance.

For example, the Ness region has a much higher area/property than any other region. Distance is highly positively correlated with travel time. Therefore using area to scale would remove a significant driver of travel costs and hence removes the operational difference that Scottish Water is aiming to measure in this special factor.

• Not a measure of scale in the econometric models.

Length of

networks

• Robust data available from Annual Returns.

• The length of network is positively correlated with distance. For example, the Ness region has a much higher network/property ratio than any other region. Distance is highly positively correlated with travel time. Therefore using length of network to scale would remove a significant driver of travel costs and hence removes the operational difference that Scottish Water is aiming to measure in this special factor.

• Only used as a measure of scale in one of the econometric models

No. of incidents • An activity that generates travel. • Does not recognise non-infrastructure activity

(i.e. visits to treatment works) • Not a measure of scale or driver of costs in

the econometric models.

No. of employees • Travel time per employee is an easy

statistic to understand. • Robust data available from Annual

Returns.

• Does not take into account differing treatment processes that require different levels of manual intervention.

• Each region employs a different balance of employees across the functions therefore we would not be comparing like with like.

• Not a measure of scale in the econometric models.


56

TRAVEL

Annex 13: Average Speed

The average speed is used within the claim to estimate the time spent travelling for operational staff

that use their own vehicles.

Average speed data from three different sources (RAC route planner, AA route planner and the

Department for Transport (DFT)) have been analysed and the results are presented in the tables below.

Specific journeys between Scottish Water offices have been run through the RAC and AA systems. The

DFT analysis is based on average speeds across all road types in England.

When considering low traffic (off-peak) conditions, all three sources produce similar results for average

speeds. The average speeds are 46 mph, 46 mph and 48 mph for RAC, AA and DFT respectively when

plotting Scottish Water journeys through their respective systems.

It is assumed that average speed is a function of traffic density which is determined by the time of day

that travel is undertaken. Car traffic is significantly higher during the peak hours of the morning and the

afternoon13. When choosing an average speed, medium traffic conditions are the most appropriate to

take account of travel during busy periods as well as off-peak conditions.

The RAC route planner is the only data source that provides information for medium traffic conditions.

Since all three sources produce similar results for low traffic conditions, it is assumed that the RAC data

is an adequate source. It is therefore used to determine the average speed in medium traffic conditions

for the report. The average speed for medium traffic conditions from the RAC route planner is 42 mph.

13 DFT (2007), ‘Road Statistics 2006: Traffic, Speeds and Congestion’


57

TRAVEL

Average speed between office locations (From RAC route planner)

Medium Traffic Low Traffic High Traffic

Journeys between main offices greater than 80 miles apart

distance (miles)

journey time (mins)

speed (mph)

journey time (mins)

speed (mph)

journey time (mins)

speed (mph)

Balmore Road (Glasgow) to Torridon House (Inverness)

168 251 40.2 228 44.2 280 36.0

Balmore Road Glasgow) to Kingshill House (Aberdeen)

151 227 39.9 205 44.2 253 35.8

Fairmilehead (Edinburgh) to Torridon House (Inverness)

159 228 41.8 207 46.1 254 37.6

Fairmilehead (Edinburgh) to Kingshill House (Aberdeen)

136 197 41.4 179 45.6 220 37.1

Castle House (Dunfermline) to Torridon House (Inverness)

141 200 42.3 182 46.5 223 37.9

Castle House (Dunfermline) to Kingshill House (Aberdeen)

118 169 41.9 153 46.3 189 37.5

Castle House (Dunfermline) to Marchmont House (Dumfries)

93 133 42.0 121 46.1 148 37.7

Bullion House (Dundee) to Torridon House (Inverness)

130 198 39.4 180 43.3 220 35.5

Bullion House (Dundee) to Marchmont House (Dumfries)

140 180 46.7 163 51.5 200 42.0

Watermark House (Livingston) to Torridon House (Inverness)

161 229 42.2 208 46.4 255 37.9

Watermark House (Livingston) to Kingshill House (Aberdeen)

138 198 41.8 179 46.3 221 37.5

Juniper House (Edinburgh) to Torridon House (Inverness)

154 219 42.2 199 46.4 244 37.9

Juniper House (Edinburgh) to Kingshill House (Aberdeen)

131 188 41.8 170 46.2 210 37.4

Marchmont House (Dumfries) to Torridon House (Inverness)

230 320 43.1 291 47.4 356 38.8

Marchmont House (Dumfries) to Kingshill House (Aberdeen)

213 295 43.3 268 47.7 329 38.8

Kingshill House (Aberdeen) to Torridon House (Inverness)

107 171 37.5 155 41.4 190 33.8


58

TRAVEL

Annex 14: Re-charge Rates for Operational Travel

The hourly rate or staff cost used for the calculation is based on 1,345 operational staff who undertake a

large amount of travelling as part of their job.

E & M Craftsman Tanker Driver

E & M Maintenance Water Operative

Maintenance Operative Wastewater Operative

Networks NSO & Inspector Modern Apprentice

The average rate is weighted by the number of employees under each of the job titles outlined above to

give an average charge out rate of £27.89 per hour. This includes overheads.

Scottish Water: Second Draft Business Plan. Complex Water Treatment

59

COMPLEX WATER TREATMENT

10. Complex Water Treatment

Executive Summary

Based upon a comparison of the proportion of distribution input (D.I.) that has been treated using the

most complex methods reported in June Returns, the Commission believes that Scottish Water incurs

lower costs in treating water to drinking water standards14. Scottish Water treats a relatively low

percentage of its D.I. using W4 level processes (see Annex 15 for details of the various categories of

treatment complexity). We understand that the Commission considers this represents a difference in

scope of activities between Scottish Water and companies in England and Wales (E&W).

We have found no evidence to suggest that the treatment processes defined as W4 are significantly more

expensive than those defined as W3. The difference in operating cost appears to be between the two

most complex categories of treatment and the three least complex categories of treatment. A

comparable proportion of Scottish Water’s D.I. has been treated using the two most complex categories

as the D.I. of companies in E&W and therefore no scope adjustment is necessary.

Basis for Scope Adjustment

Background

The level of complexity of treatment ranges from simple disinfection to W4 treatment. The various

categories of treatment complexity are summarised below (further detail is provided in Annex 15):

SD: Works providing simple disinfection only

W1: Simple disinfection plus simple physical treatment only

W2: Single stage complex physical or chemical treatment

W3: More than one stage of complex treatment but excluding processes in W4

W4: This category is intended to capture processes with very high operating costs

The Commission’s basis for the scope adjustment results from the fact that a lower proportion of Scottish

Water’s D.I. when compared with companies in E&W has received the most complex level of treatment

(W4), which involves processes with high operating costs. As a result, the Commission believes that

Scottish Water incurs lower costs in treating water to drinking water standards and considers that this

represents a difference in scope of activities between Scottish Water and companies in E&W.

14 Joint Activity Meeting – Working Group 3 (Feedback on Special Factors) held on 17/09/2008. “WICS outline that Scottish Water does not have to treat its water through expensive processes to a comparable degree of an average company in E&W. This is evident when comparing distribution input treated by W4 processes.”


60


Comparison of the Percentage of D.I. Treated using Varying Levels of Complexity

Percentage of D.I. treated by level of complexity

Company Simple

Disinfection W1 W2 W3 W4

Sutton & E Surrey 0 0 0 3 97

Dee Valley 5 0 0 4 92

Thames 0 0 4 9 87

B&W Hants 7 1 0 6 86

Three Valleys 6 0 10 14 70

Bristol 1 0 0 30 69

Anglian 2 0 15 14 68

South Staffs 5 0 22 8 66

Folkestone & Dover 28 0 7 0 64

Northumbrian 0 0 4 32 63

Southern 3 0 9 26 61

South East 3 6 17 15 59

South West 1 0 1 44 54

Severn Trent 5 0 7 38 50

United Utilities 0 0 1 61 38

Yorkshire 0 0 8 55 38

Dwr Cymru 3 1 7 53 37

Mid Kent 21 0 40 13 26

Scottish Water 2 0 12 68 18

Wessex 24 5 25 29 17

Cambridge 3 0 18 62 16

Tendring Hundred 0 0 0 84 16

Portsmouth 8 0 38 54 0

Average 3 0 8 36 52

Note: Data taken from Annual Returns 2008

The table above shows that although a lower proportion of Scottish Water’s distribution input (D.I.) than

average has been treated using the most complex treatment, it is not the lowest. A higher proportion of

Scottish Water’s D.I. has been treated using W3 level processes and a relatively low percentage using W4

level processes.

Other water companies also benefit from better raw water quality yet Ofwat does not apply a scope

adjustment. Two of these companies are considered by Ofwat to be amongst the most efficient in terms

of water service operating costs. Therefore, adjusting Scottish Water’s costs for apparently beneficial

raw water quality is not consistent with other companies who have similar operational circumstances. If

the scope adjustment is applied, Scottish Water’s efficiency position would be distorted against other

companies with comparable levels of treatment complexity.


61


Percentage of D.I. treated at W3 or W4 level

Company W3 & W4 Simple disinfection,

W1 & W2

Sutton & E Surrey 100% 0%

Tendring Hundred 100% 0%

Bristol 99% 1%

United Utilities 99% 1%

South West 98% 2%

Thames 96% 4%

Dee Valley 96% 4%

Northumbrian 96% 4%

Yorkshire 92% 8%

B&W Hants 92% 9%

Dwr Cymru 90% 10%

Severn Trent 88% 12%

Southern 88% 12%

Scottish Water 86% 14%

Three Valleys 83% 17%

Anglian 83% 17%

Cambridge 78% 22%

South East 74% 26%

South Staffs 74% 27%

Folkestone & Dover 64% 36%

Portsmouth 54% 46%

Wessex 46% 54%

Mid Kent 39% 61%

Note: Data taken from Annual Returns 2008

Furthermore, the above table shows the percentage of D.I. treated at the two most complex levels of

treatment. Scottish Water is towards the middle of the range. This indicates that some companies treat

a higher percentage of D.I. using more simple form of treatment than Scottish Water.

Quantification

Determining the incremental costs of W4 treatment

Companies in E&W do not publish breakdowns of their total operational cost by process type. Without

this data it is not possible to directly calculate the unit cost for each level of treatment. Instead,

Scottish Water used an econometric approach that is similar to that used in the efficiency models to

estimate the unit cost for separate treatment levels. The results of this analysis suggest that the unit

cost for treatment level W3 and W4 are not statistically different. The detail of this analysis is provided

in Annex 16.

The table above shows that Scottish Water treats a comparable proportion of D.I. using processes of

complexity category W3 and W4 as E&W. Given that the unit costs for both categories are not materially

different, no scope adjustment should be applied.


62


Conclusion

From our analysis, it appears that the cost of treating D.I. using the most complex processes (W4) is not

materially different from the second most complex processes (W3). Scottish Water treats a comparable

proportion of D.I. using processes of complexity category W3 and W4 as E&W. Therefore no scope

adjustment should be applied.


63


Annex 15: The Commission’s definitions of Process type, Annual Return

2008

Categories of treatment process types Examples

SD: Works providing simple disinfection only; • Marginal chlorination

W1: Simple disinfection plus simple physical treatment only;

• Rapid gravity filtration • Slow sand filtration • Pressure filtration

W2: Single stage complex physical or chemical treatment;

W3: More than one stage of complex treatment; but excluding processes in W4.

• Super chlorination • Coagulation • Flocculation • Biofiltration • pH correction • Orthophosphate dosing • Softening • Membrane filtration

W4: This category is intended to capture processes with very high operating costs;

• Ozone addition • Activated carbon / pesticide removal • UV treatment • Arsenic removal • Nitrate removal

Note: These definitions are the same as those used by Ofwat


64


Annex 16: Technical detail behind the model

Scottish Water has formulated an econometric model using available data from water companies across

England, Wales and Scotland to determine whether there is a relationship between W4 and higher costs

of treatment. The analysis indicates that the costs of W4 treatment are not significantly different from

W3, but both are significantly different from other treatment levels. This is illustrated by the model

presented below.

Model to estimate the cost differential of complex treatment:

R&T Expenditure = a * D.I. from other treatment + b * D.I. from W3 + c * D.I. from W4

(excl. EA charges & power)

‘DI from other treatment’ covers standard disinfection, W1 and W2 process types

Expenditure is in £m

DI is in Ml/day

Coefficients

(or £m/Ml/day)

Standard Error

t Stat P-value Lower 95%

Upper 95%

D.I. other 0.008 0.017 0.499 0.623 -0.026 0.043

D.I. W3 0.024 0.004 6.074 0.000 0.015 0.032

D.I. W4 0.020 0.003 7.855 0.000 0.015 0.026

The model is a simple linear one, estimated using ordinary least squares regression. This model has no

fixed costs (the intercept term is set to be zero). The coefficients estimate the cost of treatment per

Ml/day, in £m.

Whilst the coefficient on ‘D.I. from other treatment’ is not significant, it is clearly unreasonable to

expect these treatment levels to have a non-positive cost, so this term is retained in the model. This

model has an R2 value of approximately 0.9, which is very high for a cross-sectional model.

This model suggests that the cost of treatment at the W3 level of complexity is not significantly different

from the cost of treatment at the W4 level.

It should be noted that Scottish Water is not proposing this model as an alternative to the existing

resources and treatment model defined by Ofwat. It is presented here simply as a way of estimating the

cost differences between treatment levels.

Scottish Water: Second Draft Business Plan. Household Metering

65

HOUSEHOLD METERING

11. Household Metering

Executive Summary

The only material difference in the scope of activities between Scotland and England and Wales (E&W) is

that Scottish Water’s household metering penetration is low relative to that in E&W. We estimate the

value of the scope adjustment for metering to be £6.3m in 2007/08 prices.

Basis of the Scope Adjustment

Background

In E&W the weighted average level of household metering penetration is 33% for water and 32% for

sewage. In Scotland, household metering is negligible.

Econometric Models

The operating cost of household metering is included within the water and wastewater business activities

models. These models will therefore assume an operational metering cost for Scottish Water based on

that incurred per property by water and sewerage companies in E&W. Therefore it is appropriate to

remove this cost when benchmarking Scottish Water against E&W.

As the models predict operating costs for the average company, we have calculated the scope

adjustment relative to the average level of metering in E&W. This approach is in line with special factor

methodologies where claims are quantified with reference to the average in E&W. This ensures

consistency of approach across special factors and scope adjustments.

Quantification

The scope adjustment is quantified by estimating the expected cost to Scottish Water of metering the

same proportion of households as the average in E&W. For household metering, the number of additional

meters that Scottish Water would need to install to be at a similar level of penetration to the English and

Welsh average is multiplied by £4.26 per service (2007/08 prices). This is Ofwat’s projected differential

for meter reading, billing and account management compared to non-metered customers15.

The table below illustrates the additional operating cost to Scottish Water of metering a similar

proportion of properties as E&W over 2007/08 Scottish Water metering levels. The calculations are

based on Scottish Water’s total number of billed properties for water and sewerage.

15 A differential of £8.52 covering both water and sewerage services, is taken from Ofwat’s letter to Regulatory Directors - 22 September 2008 (RD 18/08). The differential is Ofwat’s means of checking that the difference between measured and unmeasured tariffs is no more than the cost of providing a metered service.

Scottish Water: Second Draft Business Plan. Household Metering

66

HOUSEHOLD METERING

Scottish Water’s additional cost if metering penetration was equivalent to E&W 2007/08

Average penetration in

E&W (2007/08)

Equivalent number of

SW properties

Current number of SW customers

metered

Annual opex (£) (reading, billing & management)

Total Cost (£m)

(1) (2) (3) (4) (5) = (2-3) x 4

Water 33% 758,338 530 4.26 3.2

Sewage 32% 724,795 180 4.26 3.1

Total 6.3

Conclusion

Scottish Water’s household metering penetration is limited relative to that in E&W. Scottish Water

estimates the value of the scope adjustment is £6.3m in 2007/08 prices.

operating cost special factors – contents - scottish water us/files/second draft business... ·...

Documents