technical annex: analysis of drought resilience · the analysis also shows that to ensure supply...

National Infrastructure Assessment | Technical Annex

1

Technical annex: Analysis of drought

resilience

July 2018


2

The size of the problem

Water is essential to people, the economy and the environment. In England,1 water

abstracted from rivers and aquifers is regulated by the Environment Agency. The vast

majority of freshwater abstracted in England is used to produce drinking water and for

industry (figure 1).

Figure 1: Freshwater use in England2

Almost half (47 per cent) of the abstraction goes to the public water supply. Water UK in

20163 highlighted the challenge of meeting public demand for water during periods of

low rainfall. These drought events are increasing in frequency and severity due to climate

change, with population growth adding to the challenge. Other abstractors also

contribute to the pressure on water resources, although to a lesser extent. Whilst the

energy sector accounts for 35 per cent of the freshwater abstraction, most of this (95 per

cent) is used for hydropower generation, thus it is not taken away from the environment.

Water demand for other types of energy generation would increase only if there is a

substantial uptake of carbon capture and storage,4 and even in this case it would only

result in low volumes (a few percentage points) compared with total freshwater use.5

The Environment Agency is tasked with ensuring that there is enough water to sustain

the environment and the life of waterbodies, supporting water quality and recharge of

aquifers. The Environment Agency revises abstraction limits periodically, issuing

“sustainability reductions” where necessary. Currently the ecosystems of at least one in

10 rivers and more than a third of groundwater bodies in England are under pressure due

to water abstraction.6

The UK Climate Change Risk Assessment 20177 identified a risk to industry from

abstraction reform and reduced water availability. This would only materialise if public

demand for water is met by increasing abstraction. Managing public demand and


3

creating additional resources to supply water even in periods of droughts, whilst

maintaining sustainable abstraction limits, should ensure that there is also sufficient

water for industry, as well as for the environment. Thus this analysis focuses on public

water supply, starting with an independent assessment of the size of the problem.

The Commission calculated future water balances under a range of droughts using the

National Infrastructure Systems Model (NISMOD),8 developed by the Infrastructure

Transitions Research Consortium. The analysis assumed no further action beyond those

listed in the previous round of Water Resources Management Plans (2014). The baseline

demand was assumed to be in line with Water UK’s “Business as Usual” scenario, under

different scenarios of population growth, climate change and drought.

• Population growth o Low – ONS 2014-based low migration population projection o High – ONS 2014-based high fertility population projection

• Climate Change o Central – medium emission Future Flows,9 average water balance scenario o Dry – medium emissions Future Flows, with less water in the South East

• Drought – drought of different probabilities of occurrence were simulated into the two Future Flows scenarios by the Water UK Long Term Planning Framework project.

o 1 per cent annual chance, corresponding to 1 in 4 probability of occurrence by 2050

o 0.5 per cent, corresponding to 1 in 7 probability of occurrence by 2050 o 0.2 per cent annual chance or 1 in 17 probability of occurrence by 2050

The above variables were combined to calculate the supply-demand balance at a

company, regional and national scale in England to look at the widest range of results.


4

Figure 2: Additional water capacity needed in England in case of drought under population and climate

scenarios10

Six water companies, serving almost 40 per cent of the English population, would

experience water deficits during a drought that has a one in four chance of occurring at

least once between now and 2050, and ten companies (serving almost 60 per cent of

households) during a drought with a one in seven chance of occurring between now and

2050 (figure 2).

Water companies are required to plan for droughts, but these include imposing

emergency restrictions – effectively cutting-off supplies to homes and businesses –

which are unlikely to be publicly or politically acceptable. It is more likely that emergency

action would be taken to sustain near normal supplies for as long as possible. This might

include tankering water across the country and removing unsustainable amounts of

water from the environment. Most options would incur very high costs and some would

result in severe environmental damage and risks to public health.

The Commission calculated the capacity needed to provide water to supply households in

periods of drought using the NISMOD model. The capacity calculated represents the

additional volume of water needed in each company to respond to drier conditions,

beyond that already available within a company (i.e. assuming that internal transfers and

investments to maintain or enhance existing capacity take place). It is also assumed that,

during these events, some additional capacity is provided by measures that reduce

demand but do not restrict essential household water use, such as hosepipe bans and

restrictions to some businesses. The calculated capacity needed accounts for

interventions in place up to 2020, thus includes those identified in the previous round of


5

Water Resource Management Plans (2014), but excludes additional interventions

proposed in the latest draft plans.

In the previous planning cycle companies assessed sufficient water to maintain

household supplies during an event comparable with the worst drought experienced by

the company. This “worst historic drought” roughly corresponds to an event with a 1 per

cent annual chance of occurring. Maintaining this existing level of resilience to 2050, in

the face of population and climate pressures, would require additional demand

management and supply for 2,700-3,000 Ml/day (depending on climate and population

scenarios).

Over and above this, the Commission estimated that England could face a shortage of

between 600 and 800 Ml/day in a severe drought with a 0.5 per cent probability and

between 800 and 1,100 Ml/day in a more extreme drought with a 0.2 per cent probability

(figure 3).

Figure 3: Water capacity needed11


6

Establishing an appropriate level of drought

resilience

To establish the appropriate level of resilience for England, the cost of providing new

infrastructure and of reducing water demand and leakage (the “resilience cost”) has been

compared to the cost of deploying emergency drought interventions.

The Commission calculated the cost of emergency interventions based on analysis by

Atkins.12 The analysis estimated the costs of supplying water during drought to avoid

imposing emergency restrictions to businesses and households on essential use (i.e. rota

cuts). It was assumed that every water company is resilient, and will maintain its

resilience, to a drought with 1 per cent annual chance of occurrence. Thus, the costs were

calculated as marginal costs compared to a 1 per cent drought. The total costs between

2020 and 2050 of implementing emergency measures to provide household water supply

during a 0.5 per cent drought, weighted by the occurrence probability, range between

£13 and £16 billion, depending on the assumed climate and population growth (figure 4).

The total costs over the same period of implementing emergency measures against a 0.2

per cent drought range between £21 billion and £27 billion.

Figure 4: Costs for the period 2020-2050 of supplying emergency measures to provide household water

supply during a drought13

Note: Costs are on a present value basis (2018 prices) weighted by the occurrence probability


7

Figure 5: Proportion of the total capacity provided by different emergency measures and corresponding

costs14

The analysis also shows that to ensure supply during drought, some costs must be borne

in advance of any event occurring. These include the provision of basic connection

infrastructure that cannot be constructed in the short timeframe of a drought. On the

other hand, extended drought permits can help tackle the deficit during mild drought,

but create risks to the environment and might reduce the availability of water to

industry. The costs of responding to a mild drought through emergency measures are

thus lower when the deficit is met mainly through cheaper but potentially higher-impact

measures. The costs increase steeply with the need for more permanent infrastructure to

meet the deficit quickly, such as connecting pipes to transfer additional abstracted water

(“Emergency abstraction and transfer” in figure 5) or emergency desalination plants.

These interventions make responding to a more extreme drought very expensive which

explains why, despite the lower likelihood of a more extreme (0.2 per cent) drought

occurring, the weighted present value costs are considerably higher.

The short-term emergency costs of providing water during a drought, weighted by their

probability of occurrence in the 2020 to 2050 period, are directly comparable with the

whole-life costs of building long-term resilience to an equivalent event. Figure 6 shows

the comparison between these two costs, including those of maintaining the current

level of drought resilience through proactive long-term measures to manage demand

and provide additional supply through infrastructure.15


8

The results show that at a national level, the cost of responding to a drought emergency

are consistently higher than those of building long-term resilience to the same event

(figure 6).

Figure 6: Comparison between emergency costs and resilience costs16

A twin track approach to tackling the risk of drought

This section describes the analysis of demand management (reducing consumption and

leakage) and infrastructure options to balance costs, benefits and risks.

Reducing consumption

Increasing the water efficiency of appliances can save considerable amounts of water.

For example, modern dual flush toilets use about half of the water compared with

traditional ones, standard showers use about one third the water of a bath, and aerated

shower heads further reduce water use.17 Behaviours are also important, for example

showering for one minute less each day can save about 3,000 litres of water per year

saving £7 on energy and £12 on water bills.18 Campaigning and public engagement also

play an important role19 and water labelling would allow consumers to make informed

decisions.20

Current efficiency initiatives are likely to result in savings of about 400 Ml/day by 205021

and new technology would increase this to 600 Ml/day over the same period, in line with

the Commission’s “central technology” scenario.22 There is strong evidence that charging

by volume leads to more efficient water use. Standard meters can reduce average

consumption by 15 per cent and smart meters23 by 17 per cent.24 Smart meters also

enable better identification of leaks, help customers understand their consumption, and

allow companies to quickly identify and target those struggling to pay their bills.25

Water companies can only impose volume based charges for new homes or occupiers,

where households use large quantities of water (e.g. power showers or swimming pools)

or in areas classified as seriously water stressed. Despite the constraints, companies are

increasing metering and bills for unmetered customers would go up. Three water


9

companies out of 18 should have near universal metering by 2030, and a further two by

2035.

Universal metering would reduce average water bills but some customers would end up

paying more than they do now. Large families may be worse off with a meter26 but this is

consistent with the fact that they consume more water. Universal metering by Southern

Water showed a reduction in the average water bill of £6 per year. More than half of

households likely to have a lower income saw a reduction in their bill (partly related to

reductions in consumption). However the average (mean) bill for households likely to

have a lower income rose by around £10 per year. This implies that losses for those

households that did pay more outweighed savings among the households that paid less,

even though there were more of the latter group.27 Assistance for lower income

households that might be worse off with metering is therefore likely to be most effective

if it is well targeted. 28

Water companies have a statutory duty to assist vulnerable customers.29 Smart metering

can help companies identify households with the highest water consumption, who might

struggle to pay their bills. Smart meters could also enable variable tariffs (recognised in

the energy sector as helpful for vulnerable consumers30) and more regular and

transparent billing (which helps households to budget31).

Overall, water bills are not seen as burdensome by customers and stakeholder

discussions indicate a generally positive attitude toward metering as observed by

Consumer Council for Water research. Companies will therefore need to work with their

customers and support them when rolling out compulsory metering.32

Commission analysis of the potential benefits of metering compared a baseline of

continuing at the current rate of meter roll-out with near universal conventional and

smart metering by 2030 and 2035. The total amount of water that would be saved in each

year ranges from 400 to 800 Ml/day (figure 7).

Figure 7: Water saved in 2050 under different metering options33

Figure 8 shows the total and marginal costs and benefits of these options. Costs include

installation, operation, replacement and carbon costs. Benefits include avoided energy

(from treating and pumping as well as household energy use) and the avoided cost of

infrastructure. These results suggest that, if the wider benefits are considered, quicker

and more comprehensive smart metering should result in savings and is at worst cost

neutral.


10

Figure 8: Costs and benefits of metering policies34

There is also evidence that a faster and better planned transition to universal metering

could unlock efficiencies and allow for more extensive engagement to help prepare

customers.35 Systematic metering should also help to identify and address water

leakage,36 target financial assistance at those households most in need and provide

benefits in all regions in England regardless of the level of water stress.

Increasing efficiency savings to 600 Ml/day by 2050 and near universal smart metering

would reduce average (measured and unmeasured) water consumption in England from

the current 141 to 118 litres per person per day, similar to Water UK’s most ambitious

(“extended” and “enhanced”) pathways.37

Leakage

About 20 per cent of the water abstracted from the environment is lost through leakage.

Water companies reduced leakage considerably in the late 1990s, but since 2000 levels

have stabilised, possibly because decisions were based on a “sustainable economic level

of leakage”. For Price Review 2019, Ofwat has changed the approach, requiring water

companies to consider reduction of at least 15 per cent from the 2020 level, or to the

level of the best performing companies (upper quartile, in terms of litres/person/day).

There are financial incentives to encourage water companies to reduce leakage.38

The Consumer Council for Water reports that leakage is one of the highest concerns for

customers,39 and that companies’ performance in managing leakage can have a big

impact on their attitude to water saving, as well as their perceptions of water companies.

However, reducing leakage levels is expensive, and fewer than a third of the water

companies have included a 15 per cent leakage reduction by 2025 in their draft planning

tables.40


11

Commission analysis considered the cost effectiveness of different leakage reduction

levels. The costs of leakage reduction are uncertain, so the Commission used ‘high’ and

‘low’ estimates based on research by Water UK and UK Water Industry Research, the

water industry’s research body.41

These costs were compared with those of providing additional infrastructure to achieve

the same level of drought resilience. Figure 9 shows the total costs of providing

resilience to a 0.5 per cent probability drought, combining different levels of leakage

reduction with additional supply infrastructure and enhanced efficiency and demand

reduction (proxied by the cost of extending metering). Additional benefits from leakage

reduction, in particular environmental benefits from reduced abstractions, can be

substantial but are not quantified in this analysis.

Figure 9: Comparison of the costs of achieving resilience to a 0.5 per cent drought including different

leakage reduction policies42

Supply infrastructure

To meet the Commission’s recommendations, 1,300 Ml/day of additional supply

infrastructure would be needed.

A range of different types of infrastructure can be used to increase water supply and

factors such as the volume of water needed, versatility, cost and environmental impact

influence the choice:

•reservoirs have significant capital costs and are generally most cost-effective when

large volumes of water are needed. They can also bring environmental benefits

(providing habitats for birds and aquatic species), as well as recreational benefits.

However, they take up large land areas and can disrupt local communities, especially

during construction. Reservoirs must be planned well before they are needed, as it

takes around ten years from the decision to build to being able to use the water

supplied

•transfers can move water from areas with surplus to those where it is needed, using

pipeline and pumping stations. In some cases existing infrastructure, rivers or canals

could be used to move water. Costs depend on the distance and topography: long or

complex transfers can be energy intensive although Victorian transfers still supply

Birmingham and Liverpool from Wales using gravity. There are risks from


12

contamination by pollutants, algae, pathogens or invasive species. A transfer network

would also allow other assets, including reservoirs, to be built further away from the

areas of highest demand, where land may be more easily available

•other options to store water, such as aquifer and surface water storages, are usually

less capital intensive but each scheme can only provide a limited volume of water

•additional water supply can also be obtained by treating non-potable sources,

including sea and waste water. Desalination has the advantage of an effectively

unlimited resource, but is very energy intensive and produces highly polluting waste.

The potential for re-use (treating waste water to a potable level) is limited by the

availability of suitable waste water and public acceptability, but it is less energy

intensive than desalination

The best approach is likely to involve a combination of these options and the industry is

well placed to determine the exact mix. The exception is water transfers. A range of

studies have all found a positive cost-benefit case for greater transfers and water

trading. 43 However, transfers currently only make up a small proportion of total supply

(about 4 per cent). This is likely because the incentives in the current system make a

strategic transfer network difficult, meaning that the decision needs to be made at a

different level.

The Commission modelled two different mixes of water supply infrastructure44:

• storage (i.e. non-transfer) infrastructure alone

• a mix of infrastructure in which transfers are used as far as practical and the

remaining capacity is provided through storage infrastructure

Although precise costs are uncertain, the costs of a combination of a network of

transfers, making up one-third to half of the resources needed, with storage

infrastructure are comparable with those of non-transfer infrastructure (figure 10).

Figure 10: Cost of supplying water via transfers and storage infrastructure vs storage alone45


13

Assessing costs and benefits of smart metering

This section sets out further details of the analysis underpinning Figure 8, which assessed

the potential costs and benefits of more widespread and faster rollout of conventional

and smart meters.

Types of meter

In this analysis, ‘smart meters’ is used to denote Advanced Metering Infrastructure

(AMI). This enables meters to be read automatically using a fixed telecoms network.

Consumption data can be made available to companies and customers in ‘real time’.

Automated Meter Reading (AMR) doesn’t benefit from a fixed network, so water

consumption data is read periodically using ‘drive-by’ technology to provide the company

with retrospective statistics (which can comprise daily, weekly, or monthly figures) used

for billing (typically twice a year).

It is assumed that meters in place by 2019 are non-smart (‘dumb’). This is fully consistent

with the situation of most water companies, and a reasonable approximation of the few

companies that are already rolling out smart meters:

• The only water company systematically rolling out smart (AMI) meters is Thames

Water. By 2020, about 16 per cent of metered households (or 8 per cent of total

households) in the area will have smart meters. As a result, this analysis has

slightly overestimated the costs of smart meters for Thames

• South West Water, United Utilities and Sutton and East Surrey Water had started

to install AMR devices for newly-metered households prior to 2018. Southern

Water also started to use AMR meters for some customers, but the proportion of

AMR meters for each company is not known. As such, the baseline and ‘dumb’

metering options are likely to underestimate cost and overestimate per capita

consumption (PCC) for these companies. However, the difference in PCC between

‘dumb’ and AMR is very small (less than 1 l/person/day by 2050)

Rollout

The baseline option follows Water UK’s business as usual profile i.e. it does not include

the proposals in the draft Water Resources Management Plans. Changes from the

baseline are assumed to start from 2020.

Replacement rate

In the options involving dumb meters, including the baseline, meters are replaced 15

years after installation. As the time of installation is not available for meters in place

before 2007, these meters are assumed to be replaced gradually until 2027. No

replacement of smart meters takes place before 2035, as the first meters are assumed to

be installed in 2020. Thus, for dumb meter options, one full round of replacements and a


14

second almost full round take place by 2050. For the smart meter scenarios, only one full

round of replacement takes place, with the second round only starting at the end of the

modelling horizon (2050).

Options modelled

Information from the Environment Agency was used for the period 2007-2015.46 For 2016,

data from Discover Water47 were used. All these data are provided for financial years, but

were assumed to be representative of calendar years (for example, data for 2016-17 were

used to represent 2016). For all options, numbers of new and replaced meters were

calculated for each year. The data of installation numbers for each year were used to

identify the numbers of replacements needed later in the period. The baseline metering

rate up to 2050 was calculated using Water UK’s ‘business as usual’ profile for the Long

Term Water Resource Management work.48 These data include metering rates in 2025,

2040 and 2065 and linear interpolation was used to derive the rates between these years.

The calculated average metering rate in 2050, assuming high population growth, is 81 per

cent, in line with the Water UK data. In addition to the baseline, the analysis considered

four metering options:

1. baseline. Gradual introduction of smart metering, following the current metering

rate. This scenario result in smart meters being installed in the entire metered

population of 2050 (81 per cent)

2. 95 per cent metering penetration, using dumb meters, by 2030

3. 95 per cent metering penetration, using smart meters, by 2030

4. 95 per cent metering penetration, using dumb meters, by 2035

5. 95 per cent metering penetration, using smart meters, by 2035.

The 95 per cent threshold has been selected as representative of universal metering.

Evidence differs on the exact proportion of ‘unmeterable’ households. Whilst Ofwat’s

2011 study49 assumed this to be 10 per cent, more recent studies50 assume 5 per cent or

lower. Whilst the unit cost of installing a meter in some of these properties would be

higher than the average, it is also likely that it will decrease as technology advances. As

such the assumption of a constant unit cost of meters should be reasonable.

Changes in consumption due to metering

Changes in per capita consumption (PCC) due solely to metering only were assessed to

calculate the benefits of each option.

Impact of metering on PCC

The impacts of metering on PCC were calculated as follows:

• for unmetered customers switching to meters, PCC was calculated by applying

the percentage reduction due to the type (dumb or smart) of meter (see figure


15

12) to the companies’ unmetered PCC from Discover Water.51 An average value52

from the literature on the effectiveness of metering was used, although this

might underestimate the benefits from switching to metered consumption. The

difference between unmetered and metered consumption reported in Discover

Water is considerably higher than the average percentage reduction from the

literature. This might be due to difficulties in accurately reporting unmetered

consumption but also to other factors, including selection effects whereby

households with lower consumption are more likely to opt in to metering, and

vice versa.

• for metered customers switching to smart meters, PCC was calculated by applying

the difference in percentage reduction between dumb and smart meters (see

figure 12) to the metered PCC available from Discover Water. This ensured that

the PCC for metered customers switching to smart meters was no higher than

their metered PCC.

With baseline metering, the average PCC reduces to 133 l/p/ d, against 130 l/p/d for the

scenario representing 95 per cent dumb metering by 2030 and 125 l/p/d for the scenario

assuming 95 per cent smart metering by 2030.

Water saved to 2050

The total amount of water saved in each year depends on the extent of metering as well

as population growth. The options considered in this study result in water savings of up

to 780 Ml/day to 2050 (figure 7). The results have been tested for the Commission’s high,

low and central population growth scenarios. The difference in water saved under each

option is between 0.5 per cent and 2 per cent regardless of the population scenario.

Since the analysis is not sensitive to the particular population scenario, the results below

are presented for the high population growth scenario only.

Costs and benefits

The framework for calculating costs and benefits of metering was developed by

Regulatory Economics53 in discussion with the Commission, building on previous analysis

carried out by Ofwat and by Mott MacDonald for the Commission.54

Costs were calculated for dumb and smart meters using the parameters listed in figure

12. Costs were separated into one-off costs for installation or replacement of each meter,

and ongoing costs for each year a meter is in operation. The one-off costs include

installation or replacement, the associated carbon cost and the social cost due to

disruption. The ongoing costs include operational costs and associated carbon costs. In

the smart metering options, the costs included replacing dumb meters with smart ones.

For each year, the number of meters installed or replaced was multiplied by the relevant

unit costs, whilst the total number of meters in place was multiplied by the ongoing

costs. The total costs were then added and discounted to 2020. The difference in costs of


16

smart meters compared to those of dumb meters was assumed to start declining after

2030, with the difference in unit costs halving every five years between 2030 and 2050.

Benefits included:

• benefits from not having to build new infrastructure

• cost and carbon savings from not having to treat and pump additional water

• carbon reductions at the household level from distributing and heating water.

These benefits occur when customers are metered, or when a dumb meter is changed to

a smart one.

Ofwat’s methodology for estimating benefits of metering did not include the avoided

energy cost to households.55 For this study, however, these benefits are included.

Benefits were estimated by multiplying unit values by the cumulative amount of water

saved each year.56 The benefits were then added for each year and discounted to 2020.

Benefits from better identification of leakage are not included in the analysis as these

were considered in the leakage analysis. Results are shown in figure 8.

Validation

The results were compared with those of other available studies (figure 11). Costs and

benefits are in line with those assessed by Ofwat in their 2011 study. The daily volume of

water (Ml/day) saved by the baseline option is comparable with the Water UK baseline.

However, this analysis resulted in higher costs due to different assumptions from those

used in the Water UK study (eg a mix of smart and traditional meters).

Scenarios Cost (£m, PV) Benefits (£m, PV) Water saved by 2050 (Ml/day)

Commission Baseline 3,800 2,700 400

Ofwat (2011)57 baseline actualised to 2018 3,400 2,300 Not available

Water UK BAU base DM strategy (to 2065)58 1,200 Not available 400

Commission 95 per cent metering by 2030 5,100 4,200 550

Ofwat (2011) 95 per cent by 2025 actualised to 2018

4,200 2,900 Not available

Commission 95 per cent metering by 2035 4,900 3,500 600

Ofwat 95 per cent by 2030 3,900 2,900 Not available

Commission 95 per cent smart metering by 2035

7,800 5,100 800

Water UK Enhanced DM strategy (to 2065) 4,000 Not available 850

Figure 11: Comparison with previous studies

Sensitivity analysis

The analysis included an assessment of the sensitivity of the results to changes in key

parameters suggesting that:


17

• the difference between the baseline level of metering and that assumed in each scenario is the main factor influencing the results

• the benefit calculation is sensitive to the assumed marginal cost of water

• the results are not sensitive to the population growth scenarios

• the cost benefit ratio is also sensitive to the future cost of smart meters.

Main causes of uncertainty:

• the available data on the initial installation and operational costs of smart meters are comparable but may vary substantially. The assumption in this study is that systematic, large scale metering programmes would unlock efficiencies

• wider environmental benefits associated with reduced consumption of water (including reduced abstraction) are not included in the analysis

• social benefits from being able to better identify vulnerable customers are not included

• the analysis only includes the benefit of metering to help reduce leakage on the customer side when this affects per capita consumption (mainly leaky appliances). Meters can help identify other leakage but these benefits are not included

• the PCC reduction attributed to meters varies from 8 per cent (reported by one reviewer), to more commonly reported 12-16 per cent, with some studies reporting up to 20-25 per cent. The value may decrease with time unless there is continuing customer engagement. A 15 per cent reduction was therefore assumed.

Parameter Source Values and notes

Initial metering Discover Water Estimates of the baseline metering for 2016

Past metering rate Environment Agency This was used to assess when those meters that are in place by 2016 need replacing

Baseline meter rollout Water UK (2016) BAU base scenario

Meter replacement frequency National Infrastructure Commission, based on expert consultation

15 years

Population growth National Infrastructure Commission: https://www.nic.org.uk/wp-content/uploads/Congestion-Capacity-Carbon-Modelling-Annex.pdf

Based on estimates at the water company level modelled by the Commission

Household occupancy rates Office for National Statistics (2017) 2.4

Metered household consumption Discover Water Average levels of metered household consumption for each water company

Unmetered household consumption Discover Water Average levels of consumption for unmetered households in each water company

Carbon cost assumptions Department for Business, Energy and Industrial Strategy:

https://www.gov.uk/government/collections/carbon-valuation--2#update-to-traded-carbon-values:-2017

The carbon costs were averaged over the period 2020-2050

Reduction in use because of installation of traditional meter and metered bill

Based on expert consultation. 15 per cent


18

Reduction in use because of installation of smart meter and metered bill

Based on expert consultation, and averaging values in literature59

17 per cent

Dumb and smart water meter installation, replacement and reading cost

Mott MacDonald for the Commission (2017)

Based on Anglian Water data, confidential. These costs are lower than other confidential data provided by Thames Water

Operational costs including:

• Water company O&M costs

• Customer portal running costs

• Cost of dealing with enquiries per year

Mott MacDonald for the Commission (2017)

Based on Anglian Water data, confidential

Social cost of disruption of reading and installing meters

Ofwat (2011) £2 per meter

Discount rate HM Treasury Green Book (2018) 3.5 per cent per annum, standard assumption for evaluating costs and benefits of projects over 30 years

Proportion of water saved that would otherwise have been heated

Ofwat (2011) 30 per cent of total water used by combination boilers

Energy consumption for heating water

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/128720/6923-how-much-energy-could-be-saved-by-making-small-cha.pdf

0.04 kWh/l

Cost of gas per kWh https://www.ukpower.co.uk/home_energy/tariffs-per-unit-kwh

2.8 pence per kWh

Cost per Ml of heated water Calculation from the above £1,120 M/l

Marginal/incremental cost of water

Regulatory Economics (2018) based on dWRMP feasible options

Varies by company on a £ per m3 basis

Marginal production cost of water Ofwat (2011) £0.4 per m3. Includes treatment and pumping costs.

Carbon emitted by production of 1 Ml of hot water

Ofwat (2011) 8.1 tCO2e

Carbon emitted by production of 1 Ml of cold water

Ofwat (2011) 0.344 tCO2e

Carbon embedded in meter installation

Ofwat (2011) 0.0198 tCO2e

Carbon embedded in meter reading Ofwat (2011) 0.00013 tCO2e

Carbon embedded in meter replacement

Ofwat (2011) 0.0048 tCO2e

Figure 12: Parameters and sources


19

End notes

1 This study covers England only, but includes the areas of Wales that provide water to Water Companies whose customers are mostly located in England. 2 Environment Agency (2018), ENV15 - Water abstraction tables for England 3 Water UK (2016), Water resources long-term planning framework 4 Byers et al. (2014), Electricity generation and cooling water use: UK pathways to 2050 5 Environment Agency (2015), Water supply and resilience and infrastructure 6 Brown et al. (2016), UK Climate Change Risk Assessment Evidence Report: Chapter 3, Natural Environment and Natural Assets 7 Surminski et al. (2016), UK Climate Change Risk Assessment Evidence Report: Chapter 6, Business and Industry 8 Hall et al. (2016), The Future of National Infrastructure. A System-of-Systems Approach 9 The Future Flow hydrology dataset was created by the Centre for Ecology and Hydrology (CEH) and contains different scenarios of future river flows under climate change conditions. Accessed at: https://www.ceh.ac.uk/services/future-flows-maps-and-datasets 10 Source: Commission calculations, based on data from Water UK, water companies and the Environment Agency and using the NISMOD model developed by the Infrastructure Transitions Research Consortium 11 Source: Commission calculations, based on data from Water UK, water companies and the Environment Agency and using the NISMOD model developed by the Infrastructure Transitions Research Consortium 12 Atkins for the National Infrastructure Commission (2018), Analysis of the cost of emergency response options during a drought 13 Source: Commission calculations, based on analysis by Atkins 14 Source: Commission calculations, based on analysis by Atkins 15 The cost of responding to a drought emergency are made up by the cost of maintaining 1 per cent level of resilience to 2050 plus the emergency costs only as shown in figure 4. The long-term resilience costs include costs of leakage reduction, demand management and infrastructure excluding intra-company transfers and small interventions needed to maintain existing capacity 16 Source: Commission calculations and analysis, using input from Atkins, Infrastructure Transitions Research Consortium and Regulatory Economics. The costs of resilience include: cost of smart metering, efficiency, 50 per cent leakage reduction (‘low’ estimate) and infrastructure, see below. 17 waterwise.org.uk/save-water/ 18 energysavingtrust.org.uk/home-energy-efficiency/energy-saving-quick-wins 19 Consumer Council for Water (2017), Water saving: helping customers to see the big picture 20 Water Label (2015), The European Water Label Industry Scheme 21 Water UK (2016), Water resources long-term planning framework, “business as usual” scenario. The costs to achieve this level of efficiency were obtained from the same study 22 National Infrastructure Commission (2017) Congestion, Capacity, Carbon: priorities for national infrastructure - Modelling Annex 23 Smart meters measure and feedback consumption data remotely at the chosen time interval 24 Based on expert consultation, and averaging values in literature including Sonderlund et al. (2014), Using Smart Meters for Household Water Consumption Feedback, Procedia Engineering 89, 990 – 997; Ornaghi and Tonin, The Effects of the Universal Metering Programme on Water Consumption, Welfare and Equity (accessed at: https://www.southampton.ac.uk/economics/research/discussion_papers/year/2018/1801-the-effects-of-the-universal-metering-programme-on-water-consumption.page); evidence provided by Thames Water, Anglian Water and Severn Trent water. 25 Pericly and Jenkins (2015), Smart meters and domestic water usage, FR/R0023 26 Priestley and Rutherford (2016), House of Commons Briefing Paper Number CBP06596 Water bills-affordability and support for household customers. 27 Ornaghi and Tonin, The Effects of the Universal Metering Programme on Water Consumption, Welfare and Equity (accessed at: https://www.southampton.ac.uk/economics/research/discussion_papers/year/2018/1801-the-effects-of-the-

https://www.ceh.ac.uk/services/future-flows-maps-and-datasets

http://www.waterwise.org.uk/save-water/

http://www.energysavingtrust.org.uk/home-energy-efficiency/energy-saving-quick-wins

https://www.southampton.ac.uk/economics/research/discussion_papers/year/2018/1801-the-effects-of-the-universal-metering-programme-on-water-consumption.page




20

universal-metering-programme-on-water-consumption.page) and Consumer Council for Water (2016), Beneath the surface: customers’ experience on universal metering 28 The Consumer Council for Water reports that in England and Wales, current social tariff schemes could support around half a million households, compared with the reported 3 million households that do not consider their bills to be affordable. Smart metering will allow companies to better identify these customers but it is unclear if existing measures are sufficient to help all who need it, regardless of any metering policy. 29 Water Industry (Charges) (Vulnerable Groups) Regulations 1999 30 Ofgem (2017), Distributional impact of time of use tariffs of gas and electricity markets and Consumer Council for Water 2017 and CCP (2017), Price and Behavioural Signals to Encourage Water Conservation 31 Consumer Council for Water (2016), Staying afloat: addressing customer vulnerability in the water sector 32 ccwater.org.uk/priorities/your-priorities/working-to-support-customers/compulsory-metering/ 33 Source: Commission calculations 34 Commission calculations using input Regulatory Economics, see further details at the bottom of this document 35 Walker, A. (2009), The Independent Review of Charging for Household Water and Sewage Services and Angling Trust, WWF-UK and others (2011), Fairness on Tap, making the case for metering 36 For example, Berger et al. (2016), Exploring the Energy Benefits of Advanced Water Metering, LBNL 1005988 37 For 2040, Water UK’s “business as usual” scenario forecasts per capita consumption of 129 l/person/day, 121 l/person/day are forecasted in the “extended” scenario and 116 l/person/day in the “enhanced” scenario for 2040 38 Ofwat (2018), Price Review 2019 methodology 39 Consumer Council for Water (2016), Water, water everywhere? 40 Although half of the companies declare the intention of reaching 15 per cent reduction by 2025 in their draft planning document (January 2018). 41 Water UK (2016), Water resources long-term planning framework and UKWIR (2011) Long term leakage goals, Report No 11/WM/08/44 42 Source: Commission calculations and analysis, using input from Infrastructure Transitions Research Consortium and Regulatory Economics 43 Deloitte (2015), Water trading – scope, benefits and options; Cave(2009) Independent Review of Competition and Innovation in Water Markets; Ofwat (2010) A study on the potential benefits of upstream markets in the water sector in England and Wales; Ernst and Young (2011), Changing course through water trading 45 Regulatory Economics for the National Infrastructure Commission (2018), Analysis of the costs of water resource management options to enhance drought resilience 44 The costs of supply infrastructure were calculated within the NISMOD model by using the cost curves developed by Regulatory Economics, see Regulatory Economics for the National Infrastructure Commission (2018), Analysis of the costs of water resource management options to enhance drought resilience 45 The costs are those for building the infrastructure needed to provide resilience against a 0.5 per cent drought assuming demand management measures achieving 118 l/p/day and 50 per cent leakage reduction by 2050, under high population and high climate change scenario. Source: Commission calculations and analysis, using input from Infrastructure Transitions Research Consortium and Regulatory Economics 46 Data on past metering installation date were used to assess when these meters would need to be replaced 47 discoverwater.co.uk 48 Water UK (2016), Water resources long-term planning framework. 49 Ofwat (2011), Exploring the costs and benefits of faster, more systematic water metering in England and Wales 50 Water UK (2016), Water resources long-term planning framework and Mott MacDonald for the NIC (2017) Value Analysis – Better Asset Management 51 discoverwater.co.uk


https://www.ccwater.org.uk/priorities/your-priorities/working-to-support-customers/compulsory-metering/

http://spnic.hmt.local/IS/National%20Infrastructure%20Assessment/A%20NIA%20document%20drafts/Technical%20Annexes/discoverwater.co.uk

https://www.dropbox.com/s/o5iydhdczz7sir8/WaterUK%20WRLTPF_Final%20Report_FINAL%20PUBLISHED.pdf?dl=0


https://www.nic.org.uk/wp-content/uploads/Value-Analysis-Better-Asset-Management.pdf

http://spnic.hmt.local/IS/National%20Infrastructure%20Assessment/A%20NIA%20document%20drafts/Technical%20Annexes/discoverwater.co.uk


21

52 Calculated from: Sonderlund et al. (2014), Using Smart Meters for Household Water Consumption Feedback, Procedia Engineering 89, 990 – 997; Ornaghi and Tonin, The Effects of the Universal Metering Programme on Water Consumption, Welfare and Equity; evidence provided by Thames Water, Anglian Water and Severn Trent water 53 Regulatory Economics for the National Infrastructure Commission (2018), Analysis of the Cost of Drought Resilience 54 Ofwat (2011), Exploring the costs and benefits of faster, more systematic water metering in England and Wales; Mott MacDonald for the National Infrastructure Commission (2017), Value Analysis – Better Asset Management 55 Ofwat (2011), Exploring the costs and benefits of faster, more systematic water metering in England and Wales 56 The hot water savings were only applied to 30 per cent of the water saved each year, assuming that the hot water produced in each household amounts to 30 per cent of the total water used by the household. This is the same assumption used by Ofwat (2011) 57 Ofwat (2011), Exploring the costs and benefits of faster, more systematic water metering in England and Wales 58 Water UK (2016), Water resources long-term planning framework 59 Including Sonderlund et al. (2014), Using Smart Meters for Household Water Consumption Feedback, Procedia Engineering 89, 990 – 997; Ornaghi and Tonin, The Effects of the Universal Metering Programme on Water Consumption, Welfare and Equity; evidence provided by Thames Water, Anglian Water and Severn Trent water

https://ore.exeter.ac.uk/repository/bitstream/handle/10871/21312/Using%20Smart%20Meters%20for%20Household%20Water%20Consumption%20Feedback.pdf?sequence=1

https://ore.exeter.ac.uk/repository/bitstream/handle/10871/21312/Using%20Smart%20Meters%20for%20Household%20Water%20Consumption%20Feedback.pdf?sequence=1








technical annex: analysis of drought resilience · the analysis also shows that to ensure supply...

Documents