Report Pumped Hydro Cost Modelling

PUMPED HYDRO
COST MODELLING
7 December 2018
Prepared by Hydro-Electric Corporation
ABN48 072 377 158
t/a Entura 89 Cambridge Park Drive,

Cambridge TAS 7170 Australia
Entura in Australia is certified to the latest version of ISO9001, ISO14001, and OHSAS18001.
©Entura. All rights reserved.
Entura has prepared this document for the sole use of the client and for a specific purpose, as
expressly stated in the document. Entura undertakes no duty nor accepts any responsibility to any
third party not being the intended recipient of this document. The information contained in this
document has been carefully compiled based on the client’s requirements and Entura’s experience,
having regard to the assumptions that Entura can reasonably be expected to make in accordance
with sound professional principles. Entura may also have relied on information provided by the client
and/or other parties to prepare this document, some of which may not have been verified. Subject to
the above conditions, Entura recommends this document should only be transmitted, reproduced or
disseminated in its entirety.
Pumped Hydro Cost Modelling Revision No: 1.0
ENTURA-10686B 7 December 2018
Document information
Document title Pumped Hydro Cost Modelling

Client organisation TasNetworks
Client contact Stephen Clark
ConsultDM number ENTURA-10686B
Project Manager Mark Hine-Haycock
Project number E307516 - P514470
Revision history
Revision 1.0
Revision description Final – original issue
Prepared by Nick West, Pippa Williams, 12/12/2018

Cameron Potter
Reviewed by Rik Van Der Kley 12/12/2018
Approved by Mark Hine-Haycock 12/12/2018
(name) (signature) (date)
Distributed to Stephen Clark TasNetworks 12/12/2018
(name) (organisation) (date)

Pumped Hydro Cost Modelling - Revision No: 1.0
This page is intentionally blank.

Executive summary
To inform future modelling of Australia’s National Electricity Market (NEM), better information is
needed on the cost of pumped hydro energy storage projects (PHES) across the NEM states.
TasNetworks engaged Entura to develop a cost model for PHES to inform its market modelling for
Project Marinus and with a view to sharing the conclusions of the study with the Australian Energy
Market Operator (AEMO) for use in future Integrated System Plan (ISP) modelling.
Neither capital costs (CAPEX) nor operations and maintenance costs (OPEX) for PHES in Australia are
well understood in the market. This is mainly due to the fact that no PHES have been built in
Australia for over 30 years. Around the rest of the world, however, rapid uptake of the technology
has continued during that period. As renewable penetration increases and thermal plant retirements
continue, interest in PHES in Australia grows. PHES has the potential to play a major role in firming
renewables so better understanding of both CAPEX and OPEX are required by electricity market
stakeholders.
This study draws on the experience of Entura, who has current data on the costs of developing,
implementing and operating hydropower projects around the world as well as exposure to the PHES
market in Australia. Entura has drawn on the wealth of experience within Hydro Tasmania to support
the estimation of OPEX for PHES.
A cost model has been developed for the capital costs of PHES across the NEM states. The model is
applied to data from the Australian National University’s atlas of pumped hydro in Australia to
develop over 400 “projects”, which have then been analysed to determine data on the likely cost of
hydropower project across different regions in the NEM. Results have been analysed for the “top”
25% of projects identified in the geographic information system (GIS).
This study has identified total potential across the various identified PHES regions of around
24,100MW with energy in storage of 390GWh. This can be broken down in terms of storage size:
$1.48m/MW for 6 hours storage, $1.70m/MW for 12 hours, $2.11m/MW for 24 hours storage and
$2.75m/MW for 48 hours storage. The high cost of 48 hour storage projects is mainly due to a low
number of such projects.
The study has found that the cost of PHES projects vary across regions, generally in relation to the
number of potential project sites in each region. Based on the data for 6 hour storage duration
projects, Tasmania has the cheapest opportunities at an average of about $1.2M/MW installed.
Project costs in NSW regions range from $1.4m/MW to $1.6m/MW. Victorian sites have an average
capital cost of $1.5m/MW, Queensland regions range from $1.5m/MW to $1.7m/MW and South
Australian project costs average $1.9m/MW. Project costs generally increase for increased storage
durations.
Projects linking existing, very large storages, such as the Snowy 2.0 project, are not considered in this
study. It’s likely that there are project sites across the NEM that have not been included in this study
and which may be better that those identified in this analysis.
OPEX includes the full business cost of operating and maintaining plant, as an incremental asset in an
existing portfolio. Both variable operation and maintenance costs (VOM) and fixed operation and
maintenance costs (FOM) have been analysed, although it has been found that VOM does not apply
to hydropower plant – all operations and maintenance costs should be captured as FOM.
i
FOM is dependent of various factors, including whether the project is a stand-alone project or part of
portfolio, the age of the plant and the installed capacity of the project.
This study has found that FOM should be taken as $16,000/MW/yr. Factors should be applied to
account for station age and installed capacity.
ii
Contents
1. Introduction 1
2. Capital Costs 3
2.1 Introduction 3
2.1.1 Objective 3
2.1.2 Limitations of this analysis 3
2.2 Approach to estimation 4
2.2.1 Basis of estimates 4
2.2.2 Cost Calculator 8
2.2.3 Benchmarking 10
2.3 Selection of “projects” 10
2.4 Analysis of results 14
2.4.1 Regional cost of pumped hydro energy storage projects 14
2.4.2 Cost of storage 19
3. Operation and maintenance costs 21

3.1 External analyses 21
3.2 Variable operation and maintenance costs 22
3.3 Fixed operation and maintenance costs 22
3.3.1 Cost validation 22
3.3.2 Station age 23
3.3.3 Portfolio vs individual costs 23
3.3.4 Station output effects 23
3.4 Summary of Results 23
4. References 25
Appendix A - Cost calculation for benchmarking
List of figures
Figure 2.1: Distribution of identified projects by REZ 12
Figure 2.2: Map showing REZ’s included in PHES regions 13
Figure 2.3: Indicative energy in storage of 6 hour storage projects across various PHES regions 15
Figure 2.4: Average unit cost of capacity for 6 hour storage projects across various PHES regions 15
iii
Figure 2.11: Unit cost of capacity vs energy in storage (top 25% of projects) 19
Figure 2.12: Impact of increasing storage time 20
List of tables
Table 2.1: Sample calculation inputs 9
Table 2.2: Sample cost estimate 10
Table 2.3: Capital cost estimation benchmarking 11
Table 2.4: REZ’s included in PHES regions 14
Table 2.5: 6 hour storage PHES projects across various PHES Regions 15
Table 3.1: Range of estimated FOM and VOM from various sources (2018 AUD) 21
Table 3.2: FOM and VOM adopted by AEMO in 2018 ISP 22
iv
1. Introduction
Pumped storage hydropower projects are a natural fit in an energy market with high penetration of
renewable energy as they help to maximise the use of the renewables that are subject to the
vagaries of the weather. Pumped storage provides a load when the there is a surplus of supply and
storage that can be recovered later. It also provides a reliable and immediate source of energy to
supply electricity to the market when renewable sources cannot.
The Australian Energy Market Operator (AEMO) has completed the inaugural Integrated System Plan
(ISP) for the National Electricity Market (NEM) (AEMO, 2018). AEMO, in developing the ISP, has taken
a view of pumped hydro energy storage costs, both CAPEX and OPEX, and capabilities (storage size)
which limits the attractiveness of pumped hydro and ensures that significant development of
pumped hydro capacity in the NEM is not required for many years.
Pumped hydro considered by the Battery of the Nation initiative considers storage sizes ranging from
7 to 48 hours. ISP modelling considered storage as having only 2 hours storage in the case of battery
energy storage systems and 6 hours in the case of pumped hydro. The value in larger storages is
different to that of smaller storages, which will be considered by AEMO in future revisions of the ISP,
assuming credible evidence is provided to support this change in assumptions.
This study is intended to inform market modelling with a better view of potential costs and
capabilities for pumped hydro energy storage across Australia’s National Electricity Market (NEM). It
is not intended to be representative of individual opportunities. Pumped hydro energy storages
(including conventional hydropower stations) are highly sensitive to the topography and remoteness
of the site.
The methodology adopted in this study provides a reasonable way to estimate the cost effectiveness
and capability of the better opportunities in certain regions across the NEM – however it is expected
that within those regions there will be sites which are more or less expensive than the regional prices
presented.
The report is split into two sections; a basis for estimating capital costs for pumped hydro projects,
and robust estimates of operational costs for pumped hydro projects.
1
2
2. Capital Costs
2.1 Introduction
2.1.1 Objective
The objective of this analysis of capital costs is to provide reasonable estimates for typical projects
across various regions in the NEM.
To achieve this objective a basis has been developed for the estimates using rates developed from
recent experience and detailed estimates by Contractors. A tool has been developed for estimating,
at a high level, the capital cost of any project based on just a few key inputs:
 Head;
 The slope distance between storages;
 The design discharge of the project; and
 The storage duration (the length of time at which the project can operate at full load).
2.1.2 Limitations of this analysis
Hydropower projects are bespoke by nature – no two sites are the same; “off the shelf” equipment is
not widely used. For this reason, applying unit costs to hydropower projects is an exercise that can
potentially lead to either underestimation or overestimation of capital costs.
Many variables can impact on the costs for each site. These include:
 Local topography
Topography determines the potential size of project, the available head and the length of
waterways, which are major cost drivers.
 Geological conditions of both dams / reservoirs, and intakes
For any type of dam / reservoir, geological conditions of the dam foundation and reservoir rim
are important in regards to designing the dam appropriately. The type and shape of a dam is
directly related to the specific site conditions and stability of the reservoir rim is impacted
especially for pumped storage projects where daily water level fluctuations are expected.
Without a site visit, geological review, and site-specific geotechnical investigation, the
assumed design and associated cost is very approximate and could be misleading.
 Geological conditions of waterways and power stations
New pumped hydro projects are likely to involve underground stations with waterways being
in the form of shafts and tunnels. This is a key risk that requires sub-surface geotechnical
investigations in the feasibility stage. However, a site visit associated with a desktop review,
mapping and analysis could provide a high level idea of the quantum of this risk. To account for
this risk without specific details of projects known, contingency is added to the cost estimates.
3
 Environmental and approval issues

Environmental and approval issues can be underestimated significantly. Environmental and
approval constraints cannot be identified without a proper review and mapping. GIS layers for
nature conservation and cultural heritage sites have been used in the screening stage to
exclude such sites but this does not mean that any identified sites outside these zones has no
issues to proceed.
 Flood risks
Flood risks for on-river dams will have direct impact on the dam’s consequence categories
which will ultimately result in additional costs for the dam and its appurtenant structures.
2.2 Approach to estimation
In the absence of such data, high level cost estimation was undertaken for different opportunities
based on Entura’s previous experience and available international data. Summary of assumptions
relating to this high level cost estimation is as follows:
2.2.1 Basis of estimates
Development of the capital cost of a project requires estimation of the cost of various project
components, including;
 Upper and lower reservoirs;
 Waterways;
 Power station (including electromechanical equipment and sub-station;
 Access tunnels and construction adits;
 Access roads to site;
 Transmission lines;
 Miscellaneous items, such as hydromechanical equipment (gates, valves etc), monitoring
instrumentation, water balancing pumps and pipework, etc.;
 Other development costs.
The basis for estimation of these components of the works follows.
Opportunities exist to re-purpose existing hydro plant for use as pumped hydro projects. These
projects generally don’t require new dams but they can have significant impact on existing
operations while construction is occurring. Such projects are not considered in this study.
2.2.1.1 Upper and lower reservoirs
Different dam types can be used in the development of a pumped hydro project. The selection of
dam type will be informed by the topography, geology, etc. For the purposes of this study, only off-
stream dams are considered. It is assumed other dam types would cost the same or less than the off-
stream dams.
4
Off-stream dams (e.g. Turkey’s nest dams)
Off-stream dams are assumed to be constructed using homogenous rockfill with durable and
economic liners on the water side. It is acknowledged that this generic arrangement might not be the
optimal type and arrangement for some sites. However, given that these dams are to be constructed
either on top of a ridge or at the bottom of hilly areas with rockfill sources available on site, a
balanced earthmoving activity (cut and fill with compaction) is potentially the most economical
solution.
Given an assumed water level fluctuation of 15 m, a nominal freeboard of 2 m, and a maximum dead
storage depth of 3 m close to the intake area, the total height of the dam would be around 20 m on
the water side. On the external side this height varies depending on the topographical conditions.
This results in the cost of storage being a squared function. i.e double the cost gives four times the
storage.
It is assumed that an average 15m dam with 2(H):1(V) internals slope and say 1.5(H):1(V) external
slope needs to be constructed around the identified area. This is a cross sectional area of 300 m2 for
the entire length of the reservoir perimeter. Also it is assumed that the liner (including sandy layer
underneath) is required over the entire internal slope of the reservoir and the reservoir floor area.
Existing dams/reservoirs and natural lakes
Using existing dams/reservoirs and natural lakes negates the need to construct a new dam, however
connecting the new waterways to the existing storage requires a new intake.
Some stabilisation works would be required on the existing dam itself and/or the reservoir rim. This
is a major unknown factor in the cost estimation as it is very site specific and depends on the local
topography and geological and geotechnical conditions. A judgement on the need and extent of such
works can realistically be made only after initial site investigations.
Natural sinks / mining pits
Natural sinks and mining pits are assumed to have stable walls either naturally or based on the
mining requirements assuming that there is no water level fluctuation against them. Such transient
and cyclic loading, which is inevitable for a pumped hydro project can de-stabilise the pit wall
depending on the geological features, their alignments and characteristics. It is likely that pit wall
stabilisation would need to be carried out within the fluctuation zone.
A lower intake needs to be constructed in a similar fashion to those in existing dams/reservoirs.
This analysis does not have the resolution to consider mining pits for use as pumped hydro storages.
Valley dams
For valley dams (on-stream storages) a number of considerations need to be taken into account,
including:
 The foundation conditions;
 The size and position of the spillway;
 The location and size of the intake structure;
5
 Upstream and downstream dam slopes along with dam crest width;
 The alignment of the dam axis;
 Temporary diversion arrangements.
Cost estimation of the reservoir rim stabilisation and intake structure is similar to that of existing
dams.
2.2.1.2 Intakes
This study assumes that two new intakes will be required – one at both the upstream and
downstream ends of the waterway. Site specific temporary works such as a cofferdam may be
required to construct the intake.
For the purposes of this study, it is assumed that the intake is a horizontal diffuser shape that is
permanently submerged under water with adequate depth at the inlet. As an approximation, it can
be assumed that the cost increases/decreases with the square root of the discharge for such a
structure. This is based on keeping the design average velocity of the water at the intake to about
1m/s.
2.2.1.3 Waterways
For a pumped hydro project, it is normally optimal to have a layout with the shortest waterways
connecting the two reservoirs. This not only reduces the head losses in the system (thus increasing
the cyclic efficiency) but also reduces the project costs. Based on this, and taking into consideration
that fixed speed reversible units (that are often the optimal / economic arrangement) require
substantial submergence in relation to the minimum operating level of the lower reservoir;
underground power stations and waterways are the most economic solutions. There are some
occasions that surface waterways and power stations or a combination of both could be more
favourable, such as when design discharges are relatively low, so the cost of surface penstocks is
relatively low.
For the purpose of this study, fully underground arrangements have been assumed with about 25%
of the average gross head being steel lined (including the branches) and the rest concrete lined.
Other assumptions are:
 The average velocity in the largest section of the waterways is about 4 m/s
 At least two pump/turbine units are involved.
 A single shaft/tunnel arrangement is considered with no consideration of the practicalities of
installing steel liners in large diameter pressure tunnels.
2.2.1.4 Power Station (including switchyard)
Electrical and mechanical (E&M)
Estimation of E&M costs relates significantly to the head and discharge of the station as well as type
of technology adopted. It appears that for large scale pumped storage projects, fixed speed
reversible units are the most economic choice most of the time.
6
Variable speed units and ternary (units with separate pump impellers and turbine runners) on the
same shaft) sets are in the order of 30% and 50% more expensive respectively, which equates to an
increase in total project cost in the order of 10% to 15%. Depending on the situation, there may be
value in installing a combination of fixed and variable speed machines. This is outside the scope of
this study.
For the purposes of this study, fixed speed units are assumed. The cost of these units is proportional
to head and flow.
Also for the purpose of this study, the installed capacity of all projects is limited to 600MW.
Civil
Civil costs of the power station, switchyard and associated miscellaneous elements such as
cable/exhaust shaft/tunnels or routes are absolutely site specific and any estimation could be up to
30% over/under estimation. The cost of powerhouse civil works is also proportional1 to the head and
flow of the project, since the size of the machines governs the size of the powerhouse.
2.2.1.5 Access tunnel and construction adits
The cost of access tunnels is site specific as the shortest access route to the station is defined by the
topographic conditions of the site and the location of existing roads. For the purpose of this study, it
is assumed that this length is around 75% of the horizontal distance between the two reservoirs. It is
assumed that that the tunnel is 6m diameter. An allowance of 500m has been allowed for
construction adits.
2.2.1.6 Access roads
It is assumed that access to site is available for both construction and operation. A unit rate for site
access roads is adopted and applied to a length of access road that is proportional to the head of the
project. i.e. site access roads are assumed to have an average gradient of 10%, meaning for a project
with 400m head, the assumed site access road length will be about 4km.
A flat rate is adopted for required upgrade works for roads to site.
2.2.1.7 Transmission line
Transmission line costs are made up of:

 a fixed component that covers the cost of new substation works – both at the power station
and the remote connection; and
 a variable cost that accounts for the cost of a new transmission line to the nearest
transmission substation.
1
Turbine dimension is proportional to square root of flow. Therefore, if flow is doubled, the volume of
powerhouse increase roughly to 1.414x1.414x1.414 = 2.82 times. Turbine size is not affected by head.
Generator size is affected by head and flow and affected by speed. For constant speed, double power = 1.414 x
generator diameter with same length
7
Transmission line costs do not consider the cost of wider transmission augmentation. It is assumed
that this cost will not be the responsibility of the developer of the pumped hydro project.
2.2.1.8 Miscellaneous
Items such as hydraulic instrumentation, hydro-mechanical works such as gates, valves, etc. that is
not included in the power station E&M, water balancing pumps and pipework, etc. An estimation of
2% of total cost of the works excluding access roads/tunnels is adopted.
2.2.1.9 Other development Costs
The direct cost estimations above include contractor’s indirect costs and profit.
 Preliminaries and general – 20%
 Design and approvals – 10%
 Owner’s costs – 5%
 Contingency – 20%
2.2.2 Cost Calculator
Using the basis of estimates described in Section 2.2.1, a cost calculator developed for theoretical
pumped hydro projects with various physical characteristics including:
 Gross head,
 Waterway length,
 Storage Volume; and
 Storage time2
 Distance to nearest NEM substation.
From these inputs, various parameters are determined for use in the cost estimate including:
 Design discharge (m³/s);
 Installed capacity (MW);
 Energy in storage (MWh);
 Storage perimeter (m);
 Embankment volume (m³);
 Embankment liner area (m²);
 Waterway diameter (m); and
 Site access roads lengths.
From these parameters, costs are estimated for the following elements:
2
This study considers storage sizes of 6, 12, 24 and 48 hours.
8
 Upper reservoir;
 Lower reservoir;
 Intakes;
 Waterways;
 Power station (E&M);
 Power station (civil);
 Access tunnels and construction adits;
 Access roads to site;
 Transmission line;
 Miscellaneous; and
 Other development costs (preliminaries and general, design and approvals, owner’s costs,
contingency).
An example calculation from the tool is provided in Table 2.1 and Table 2.2.
Table 2.1: Sample calculation inputs

Inputs
Serial Decription Value
Input Values
1 Head, H (m) 400
2 Waterway length, L (m) 1,500
3 Storage volume (Mm³) 4.5
4 Storage time (hr) 12
5 Transmission length (km) 20
Calculated values
5 Length: Head Ratio 3.8
6 Flow (m³/s) 104
7 Pump/Turbine Efficiency3 94%
8 Motor/Generator Efficiency3 98%
9 Installed Capacity (MW) 377
10 Energy in storage (MWh) 4,518
11 Storage perimeter (m) 1,942
12 Embankment volume (m³) 912,564
13 Embankment liner area (m²) 386,832
14 Waterway diameter (m) 5.8
15 Access road length (km) 4.0
3
Use of these values gives a round trip efficiency of 85%. This calculation does not take into account headloss or
transmission losses, which would typically result in round trip efficiency of 75% to 80%.
9
Table 2.2: Sample cost estimate

Cost Estimate
Serial Description Amount
1 Upper reservoir $53,364,852
1.1 Embankment $45,628,209
1.2 Liner $7,736,643
2 Lower reservoir $53,364,852
2.1 Embankment $45,628,209
2.2 Liner $7,736,643
3 Intakes $16,137,431
4 Waterways $47,636,431
5 Power Station (E&M) $137,313,041
6 Power Station (civil) $34,328,260
7 Access tunnels and construction adits $24,000,000
8 Access roads to site $2,000,000
9 Transmission line $16,000,000
10 Miscellaneous $8,082,897
11 Sub-total works $412,227,764
12 Other development costs $267,948,046
12.1 Preliminaries and general – 20% $82,445,553
12.2 Design and approvals – 10% $41,222,776
12.3 Owner’s costs – 5% $20,611,388
12.4 Contingency – 20% $123,668,329
13 Total $680,175,810
2.2.3 Benchmarking
The cost estimation tool has been benchmarked against public figures for Australian projects
currently under development. The benchmarking is shown in Table 2.3. Calculation tables similar to
that shown in Table 2.1 are provided in Appendix A.
It should be noted that contingency is included in the cost estimates for the benchmarking exercise.
The unknowns for projects that have been subject to further analysis are likely to have been assessed
and may result in reduced contingencies being applied; hence the higher costs from the cost
calculator. Additionally, the benchmarked projects variously include or exclude the cost of
transmission connection. Transmission is therefore excluded from the costs of projects used in the
benchmarking exercise.
2.3 Selection of “projects”
One of the objectives of this study is to identify any differences in costs for pumped hydro projects
across different regions. In order to reduce the potential options to a reasonable number, regions are
defined according to AEMO’s proposed Renewable Energy Zones. It is expected the costs of
attractive projects within REZ’s will be representative of the cost of all potentially attractive projects
The reason for grouping projects within REZ’s is the potential benefit of synchronous generation
within these zones and the likelihood of sufficient transmission capacity once the REZ is
implemented.
10
Table 2.3: Capital cost estimation benchmarking
Project Location Proponent Installed Reported Estimated Estimated Estimated/

Name Capacity Unit Cost Cost Unit Cost Reported
(MW) ($m/MW) ($m)# ($m/MW)
Kidston QLD Genex 250 1.6 488 2.0 1.22
(K2-Hydro) Power
Limited
Snowy 2.0 NSW Snowy 2000 2.25 5,733 2.9 1.27
Hydro
Cultana SA Energy 225 2.1 625 2.8 1.32
Australia
Goat Hill SA Altura 230 1.8 529 2.3 1.28
Group
Highbury SA Tilt 300 1.4 628 2.1 1.50
Renewables
Kanmantoo SA Hillgrove 220 1.0 440 2.0 2.00
Resources
#
Estimated using the tool developed for use in this study
Selecting sites within each REZ was done using the ANU database of potential upper storages (Lu et
al, 2018). Potential lower storages were sited near (within 5km) of clusters of upper storages. In all,
49 lower storage sites were identified. Using GIS, the selected lower storages were paired within
upper storages within 5km to create “projects”. This methodology may result in sub-optimal design
however, for this level of study, the approach is considered appropriate.
“Projects” identified in the GIS were defined in terms of head (the difference in elevation between
the upper and lower reservoir), waterway length (the distance between the upper and lower
reservoir), storage volume (ANU upper storage volume adopted) and storage time (6 hours, 12 hours,
24 hours and 48 hours). An algorithm in the GIS then applied the cost calculator as described in
Section 2.2.
Siting of projects does not consider land tenure, geological conditions or local environmental/social
values beyond excluding sites within CAPAD designated areas. Projects considered in this study may
not be viable from a technical perspective. Further, there are likely to be many project sites outside
REZ’s.
In total, 410 “projects” were identified across 12 REZ’s. The distribution of projects is shown in
Figure 2.1.
REZ’s are grouped to enable regional statistics to be provided for PHES projects. PHES regions and
the REZ’s included in those regions are shown in Figure 2.2 and described in Table 2.4. REZ’s not
included in the PHES regions had no PHES projects identified for inclusion in this study.
11
Figure 2.1: Distribution of identified projects by REZ
12
Figure 2.2: Map showing REZ’s included in PHES regions
13
Table 2.4: REZ’s included in PHES regions
PHES Region Included REZ’s

North Queensland North Queensland
Isaac
South Queensland Fitzroy
Darling Downs
North New South Wales Northern New South Wales
New England
Central New South Wales Central New South Wales
Southern New South Wales Tumut
Victoria Ovens Murray
South Australia Northern South Australia
Tasmania Tasmania Midlands
North-West Tasmania
2.4 Analysis of results
2.4.1 Regional cost of pumped hydro energy storage projects
Costs are presented for the “top” 25% of projects from the identified list. These spread across the
various PHES regions. Results in terms of indicative installed capacity, energy in storage and the unit
cost of capacity for the top 25% of projects are provided in the following charts and tables. It should
be noted that the above categories are not mutually exclusive. A single project could be repeated
across multiple “energy in storage” analyses.
For 48 hour projects, the sample size is significantly smaller than for projects with shorter duration
storages. The costs of some 48 hour projects are considerably higher due to the reduced installed
capacities. Excluding projects with unit costs of over $3m/MW brings the average cost down to
$2.39m/MW, which brings the cost of 48 hour projects into line with the costs of 6, 12 and 24 hours
projects.
14
Table 2.5: 6 hour storage PHES projects across various PHES Regions
PHES Region Total capacity of Total energy in Average unit cost of

projects (MW) storage (MWh) capacity ($m/MW)
Central NSW 1,200 7,200 1.53
North NSW 600 3,600 1.36
North QLD 1,200 7,200 1.53
SA 500 2,900 1.93
South NSW 1,600 9,600 1.63
South QLD 600 3,600 1.73
TAS 2,200 13,000 1.16
VIC 1,200 6,900 1.53
Total: 9,100 Total: 54,000 Average: 1.48
6900 7200
Central NSW
North NSW
3600
North QLD
SA
13000
7200 South NSW
South QLD
TAS
2900
VIC
3600
9600
Figure 2.3: Indicative energy in storage of 6 hour storage projects across various PHES regions
2.50
2.00
1.50
1.00
0.50
0.00
Central North North SA South South TAS VIC
NSW NSW QLD NSW QLD
Figure 2.4: Average unit cost of capacity for 6 hour storage projects across various PHES regions
15

Central NSW 800 9,400 1.87
North NSW 600 7,200 1.52
North QLD 900 11,100 1.80
SA 200 2,900 2.64
South NSW 1,200 14,400 1.76
South QLD 600 7,200 1.90
TAS 2,000 23,800 1.28
VIC 1,200 13,800 1.75
9400
13800
Central NSW
7200 North NSW
North QLD
SA
11000
South NSW
23800
South QLD
2900
TAS
VIC
14400
7200
3.00
2.00
1.00
0.00
NSW NSW QLD NSW QLD
16

Central NSW 200 5,300 2.43
North NSW 600 14,400 1.85
North QLD 500 10,900 1.98
SA 200 4,500 3.75
South NSW 700 17,200 2.26
South QLD 600 14,400 2.17
TAS 1,600 37,500 1.39
VIC 700 17,400 2.25
5300
17400
14400 Central NSW
North NSW
North QLD
10900
SA
4500 South NSW
37500 South QLD
TAS
17200
VIC
14400
4.00
3.00
2.00
1.00
0.00
NSW NSW QLD NSW QLD
17

Central NSW 100 5,400 3.79
North NSW 600 28,800 2.11
North QLD 200 10,900 2.92
SA - - -
South NSW 200 9,200 3.08
South QLD 300 16,000 2.96
TAS 800 37,500 1.57
VIC 400 17,400 3.38
5300
17400
14400 Central NSW
North NSW
North QLD
10900
SA
4500 South NSW
37500 South QLD
TAS
17200
VIC
14400
4.00
3.00
2.00
1.00
0.00
NSW NSW QLD NSW QLD
18
The reasons for the differences in cost for the various regions include:
 The waterway length to head ratio varies according to the region in which projects are located
– projects with lower waterway length to head ratios are generally cheaper;
 The installed capacity for a project is a significant driver in determining the unit cost. Installed
capacity is related to head and storage size in this analysis. In regions where storages are
relatively small and head is relatively low (eg South Australia), costs are generally higher;
 Regional geology is a factor contributing to the size of storages, which is related to both the
installed capacity and the capital cost.
Figure 2.11 shows the top 25% of projects in terms of the unit cost of capacity and the energy in
storage, which gives an indication of the relative costs of larger storages. In general, the graph shows
that projects with larger storages are more expensive.
2.4.2 Cost of storage
A byproduct of the analysis is an illustration of the impact of increasing the storage time for project
with the same reservoir sizes. That is, the relative increase in the unit cost of capacity with reduced
installed capacity and reduced waterway and associated infrastructure sizes.
Figure 2.11: Unit cost of capacity vs energy in storage (top 25% of projects)
19
The main reason that unit cost of capacity increases with increasing storage time is the cost of new
storages4 and the length of waterways remain the same but there is less installed capacity. Statistics
highlighting the impact of increasing storage time are presented in Figure 2.12. On average, an
increase in storage time from 6 hours to 12 hours results in a 40% increase in the unit cost of
capacity. An increase in storage time from 6 hours to 24 hours results in a 120% increase in the unit
cost of capacity.
In a very small number of cases, there may be a decrease in the unit cost of capacity with increasing
storage time. This is only likely in the case where the cost of new dams is relatively high.
4.5
4
Relative unit cost of capacity
3.5
2.5
2
Average
1.5 90% exceeding
10% exceeding
1
0 10 20 30 40 50 60
Storage time (hours)
Figure 2.12: Impact of increasing storage time
4
It is assumed the analysis by ANU has maximised the potential volume of the upper storage. If it is possible to
build a larger storage on the same site, the cost increase will not be as significant.
20
3. Operation and maintenance costs

Operation and maintenance cost estimates were based on several US and Australian reviews and
estimates, verified with Entura expertise and leveraging Hydro Tasmania asset management
experience.
3.1 External analyses
A number of sources have been consulted to determine the operation & maintenance costs for
traditional and pumped hydro plant as shown in Table 3.1. Reputable sources vary greatly in their
estimates of costs per MW. For example, three US estimates are compared below, along with
estimates used by the Australian Energy Market Operator. Original estimates have been converted to
2018 AUD5.
Table 3.1: Range of estimated FOM and VOM from various sources (2018 AUD)
Total $/MW/year,
FOM VOM
Source assuming nominal
AUD/MW/year AUD/MWh
0.25 capacity factor
Conventional hydro
Cost and Performance Data for Power
Generation Technologies, prepared for the
22,056 8.82 41,377
National Renewable Energy Laboratory
Black & Veatch 2012
Combined Data for Conventional and Pumped Hydro
Electric Power Annual 2016
US Energy Information Administration 14.76 32,320
2017, revised 2018
Pumped Hydro
Cost and Performance Data for Power
Generation Technologies, prepared for the
45,288 - 45,288
National Renewable Energy Laboratory
Black & Veatch 2012
DOE/EPRI Electricity Storage Handbook in
Collaboration with NRECA 9,182 0.42 10,110
Sandia National Laboratories 2015
2018 Integrated System Plan modelling
assumptions 5,000 5.00 15,950
Australian Energy Market Operator 2018
5
Inflation: World Bank World Development Indicators: https://data.worldbank.org/indicator/FP.CPI.TOTL.ZG
Exchange rates: OFX: https://www.ofx.com/en-au/
21
In contrast, the assumption used for conventional hydro in the Australian Energy Market Operator’s
modelling is significantly higher as shown in Table 3.2.
Table 3.2: FOM and VOM adopted by AEMO in 2018 ISP

Total, assuming
FOM VOM
Source nominal 0.25
AUD/MW/year AUD/MWh
capacity factor
Conventional hydro
2018 Integrated System Plan modelling
assumptions 58,250 7.19 73,996
Australian Energy Market Operator 2018
The AEMO ISP assumption for O&M costs for conventional hydro generators is a significant outlier in
terms of published cost assumptions, as well as Entura’s experience.
By comparing these to Entura experience, it is surmised that these differences can be attributed to
differing scopes. However, very little detail is available regarding the scope of publicly available
estimates.
This report estimates the full business cost of operating and maintaining plant, as an incremental
asset in an existing portfolio.
3.2 Variable operation and maintenance costs
Variable operation and maintenance costs (VOM) are defined as O&M costs which are proportional
to the amount of energy generated, in MWh. Hydroelectric power stations are often unmanned, and
variable costs would be predominantly maintenance costs due to plant deterioration associated with
running. Deterioration of hydro power plant is typically age-based, exacerbated by two factors:
running outside the machine’s efficient operating range, and machine starts and stops. These factors
are both highly dependent on machine design, both in terms of the breadth of operation a generator
is designed for, and its robustness to off-spec operation. Neither of these factors is proportional to
the capacity factor of the plant. Therefore they are not accurately represented by measuring VOM. It
is therefore recommended that hydropower operation and maintenance costs are captured as Fixed
Operation and Maintenance (FOM) costs only, rather than assigning some portion of these costs to
Variable Operation and Maintenance (VOM).
3.3 Fixed operation and maintenance costs
3.3.1 Cost validation
The costs have been validated against Entura’s experience. The costs were taken to be the full cost of
the business of operating a hydroelectric generating asset, as an incremental cost to an existing
business.
It has been noted that Hydro Tasmania’s portfolio has historically been typified by plentiful capacity,
constrained by energy (water) availability. As such, the value of certain industry-standard metrics
(such as start reliability or planned outage factor) has been suppressed. In the future, if Tasmania is
more connected with the rest of the NEM, there will be more value in improving these metrics. It is
22
acknowledged that this may increase maintenance costs. However given the broad alignment
between Hydro Tasmania experience and external estimates, and the approximate nature of the cost
estimates, this potential uplift is assumed to be negligible for the purposes of this report.
3.3.2 Station age
Entura experience also indicates that costs are well correlated with station age, with stations less
than 50 years old typically having much lower operation and maintenance costs than older stations.
For stations over 50 years old, a factor of 1.5 should be applied to the original estimate.
3.3.3 Portfolio vs individual costs
It is noted that Hydro Tasmania’s portfolio consists of 27 power stations, which are managed in six
geographic groupings. The relative proximity of stations within each geographic grouping enables
sharing of O&M resources between stations. However the average size of Hydro Tasmania’s stations
is around 85MW6, so the reduced costs of operating a geographic cluster of stations is offset by the
relatively small size of each station.
It is also noted that Hydro Tasmania has a large portfolio of stations, with over 100 years of
operational experience. However it is assumed that if an owner were to invest in a single
hydropower or pumped hydro station, they would outsource operation and maintenance to a party
with more experience and economy of scale. Other business costs, such as insurance, will generally
be less overall in a portfolio of projects.
3.3.4 Station output effects
Entura’s experience also indicates that very small stations (less than 20 MW), have much higher costs
per MW (more than double the cost per MW of larger stations of similar age).
For stations with installed capacity less than 100MW, a factor of 1.5 should be applied to the original
estimate.
For stations with installed capacity less than 20MW, a factor of 4 should be applied to the original
estimate.
3.4 Summary of Results
On the basis of this, it is recommended that the AEMO ISP value is a good starting point, but should
be augmented with a number of minor improvements:
1. VOM is not meaningful for hydropower projects, as it does not take into account the most
damaging aspects of operation. As such, FOM only should be used – a single value of
$16,000/MW/yr. This is the equivalent of the existing AEMO ISP O&M cost assumption for
pumped hydro, consolidated from a FOM and VOM into a single FOM.
6
Tasmania’s hydropower system was designed to be almost the sole provider of electricity in the state.
Therefore station MW capacity is typically low compared to total storage capacity and energy generation, with
a portfolio capacity factor of around 45%.
23
2. This should be adjusted for age and size of the station (not machine).
a. For stations over 50 years old, a factor of 1.5 should be applied to the base estimate.
b. For stations with installed capacity less than 100MW, a factor of 1.5 should be
applied to the base estimate.
c. For stations with installed capacity less than 20MW, a factor of 4 should be applied
to the base estimate.
Note that this means for an old, small, station, the factor could be as large as 6.
As stated previously, this analysis assumes that the station is operated as part of a portfolio. If a
station were not part of a portfolio, it is assumed that operation and maintenance would be
outsourced.
24
4. References
AEMO, 2018, Integrated System Plan for the National Electricity Market, July
Lu, B., Stocks, M., Blakers, A., Anderson, K., 2018, Geographic information system algorithms to
locate prospective sites for pumped hydro energy storage, Applied Energy 222 (2018) 300–312
25
26
A Cost calculation for benchmarking

Note: Transmission costs excluded.
A.1 Kidston (K2-Hydro)

Input Values
1 Head, H (m) 185
2 Waterway length, L (m) 450
4 Storage time (hr) 5.7
Calculated values
6 Flow (m³/s) 150
A.2 Snowy 2.0

Input Values
1 Head, H (m) 681
4 Storage time (hr) 178
Calculated values
6 Flow (m³/s) 325
7 Installed Capacity (MW) 1,998
9 Storage perimeter (m) -
10 Embankment volume (m³) -
11 Embankment liner area (m²) -
27
A.3 Cultana

Input Values
1 Head, H (m) 260
Calculated values
6 Flow (m³/s) 96
A.4 Goat Hill

Input Values
1 Head, H (m) 200
Calculated values
6 Flow (m³/s) 127
28
A.5 Highbury

Input Values
1 Head, H (m) 156
Calculated values
6 Flow (m³/s) 213
A.6 Kanmantoo

Input Values
1 Head, H (m) 250
2 Waterway length, L (m) 650
Calculated values
6 Flow (m³/s) 98
29
30

Report Pumped Hydro Cost Modelling

Uploaded by

Copyright:

Available Formats

Report Pumped Hydro Cost Modelling

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Report Pumped Hydro Cost Modelling

Uploaded by

Copyright:

Available Formats

What is the purpose of the document?

What is the purpose of the document?

What information is provided about PHES costs?

What information is provided about PHES costs?

PUMPED HYDRO

t/a Entura 89 Cambridge Park Drive,

©Entura. All rights reserved.

Document title Pumped Hydro Cost Modelling

Revision description Final – original issue

Prepared by Nick West, Pippa Williams, 12/12/2018

Approved by Mark Hine-Haycock 12/12/2018

(name) (signature) (date)

Distributed to Stephen Clark TasNetworks 12/12/2018

(name) (organisation) (date)

This page is intentionally blank.

3. Operation and maintenance costs 21

Appendix A - Cost calculation for benchmarking

Figure 2.1: Distribution of identified projects by REZ 12

Figure 2.2: Map showing REZ’s included in PHES regions 13

Figure 2.12: Impact of increasing storage time 20

Table 2.1: Sample calculation inputs 9

Table 2.2: Sample cost estimate 10

Table 2.3: Capital cost estimation benchmarking 11

Table 2.4: REZ’s included in PHES regions 14

Table 3.2: FOM and VOM adopted by AEMO in 2018 ISP 22

This page is intentionally blank.

2.1.2 Limitations of this analysis

 Environmental and approval issues

2.2 Approach to estimation

2.2.1 Basis of estimates

The basis for estimation of these components of the works follows.

2.2.1.1 Upper and lower reservoirs

Off-stream dams (e.g. Turkey’s nest dams)

Existing dams/reservoirs and natural lakes

Natural sinks / mining pits

A lower intake needs to be constructed in a similar fashion to those in existing dams/reservoirs.

2.2.1.4 Power Station (including switchyard)

Electrical and mechanical (E&M)

2.2.1.5 Access tunnel and construction adits

2.2.1.6 Access roads

2.2.1.7 Transmission line

Transmission line costs are made up of:

2.2.1.9 Other development Costs

2.2.2 Cost Calculator

Table 2.1: Sample calculation inputs

Table 2.2: Sample cost estimate

2.3 Selection of “projects”

Table 2.3: Capital cost estimation benchmarking

Project Location Proponent Installed Reported Estimated Estimated Estimated/

Figure 2.1: Distribution of identified projects by REZ

Figure 2.2: Map showing REZ’s included in PHES regions

Table 2.4: REZ’s included in PHES regions

PHES Region Included REZ’s

2.4 Analysis of results

2.4.1 Regional cost of pumped hydro energy storage projects

PHES Region Total capacity of Total energy in Average unit cost of

PHES Region Total capacity of Total energy in Average unit cost of

PHES Region Total capacity of Total energy in Average unit cost of

PHES Region Total capacity of Total energy in Average unit cost of