WAIKATO AND UPPER NORTH ISLAND VOLTAGE MANAGEMENT INVESTIGATION UPDATE

UPDATE ON OUR ANALYSIS TO DATE

Transpower New Zealand Limited

May 2018

C O P Y R I G H T © 2 0 1 8 T R A N S P O W E R N E W Z E A L A N D L I M I T E D . A L L R I G H T S R E S E R V E D

This document is protected by copyright vested in Transpower New Zealand Limited (“Transpower”). No part of the document may be reproduced or transmitted in any form by any means including, without limitation, electronic, photocopying, recording or otherwise,

without the prior written permission of Transpower. No information embodied in the documents which is not already in the public domain

shall be communicated in any manner whatsoever to any third party without the prior written consent of Transpower.

Any breach of the above obligations may be restrained by legal proceedings seeking remedies including injunctions, damages and costs.

L I M I T A T I O N O F L I A B I L I T Y / D I S C L A I M E R O F W A R R A N T Y

This document is produced for internal use only and has not been approved for external release. Its conclusions are based on the

information currently available to Transpower and may change as further information becomes available either internally or externally.

Summary

Recent actual, and announced, decommissioning of major generation plants in the Upper

North Island – the region north of Huntly and including Auckland – represent significant

changes for the New Zealand power system and create issues with voltage management in

the Waikato and the Upper North Island, and with thermal transfer into the region.

The voltage management issues are being investigated as the Waikato and Upper North

Island Voltage Management (WUNIVM) Investigation. We expect it to lead to us submitting

a Major Capex Proposal (MCP) to the Commerce Commission to undertake investment and

as such are following an MCP process. We undertook our long-list consultation under that

process in July 20161. While we have not yet fully developed our preferred option to

address the WUNIVM need, we have made significant progress.

This document provides an update on the WUNIVM investigation to indicate to interested

parties where the WUNIVM investigation is heading and what they might expect in our

proposal. The major components of our evolving preferred option are:

Components of the evolving preferred option Expected need date

Shunt capacitors in the Waikato 2020

Series capacitors on the BHL-WKM circuits 2022

Demand management scheme to manage high impact, extremely low probability contingencies

2022

Additional components to manage voltage stability 2024+

The WUNIVM investigation is proceeding and we plan to put our preferred option to formal

short-list consultation in the first half of 2019. The evolving preferred option outlined above

may change:

• as the WUNIVM technical, reliability and economic analysis proceeds

• if any generation commitments are made at Huntly or north

• with any material revision to demand forecasts

• consequent to submissions received on our short-list consultation.

1 Available at www.transpower.co.nz/waikato-and-upper-north-island-voltage-management-investigation.

Contents

Summary ............................................................................................................................. 2

Glossary............................................................................................................................... 4

1 Introduction and background ..................................................................................... 5

Need for the WUNIVM Investigation ...................................................................... 5

Impact of Genesis Energy’s decisions on Huntly .................................................... 7

Evolving view of the preferred WUNIVM investment option .................................... 8

2 WUNIVM’s evolving preferred option ......................................................................... 9

Shunt capacitors .................................................................................................... 9

Series capacitors ................................................................................................. 10

Demand-management scheme ............................................................................ 11

Additional components to manage voltage stability .............................................. 12

3 Modelled projects .................................................................................................... 13

4 Next steps ............................................................................................................... 13

Voltage stability limits .................................................................................... 14

Figures

FIGURE 1 - WUNIVM NEED GRAPH .................................................................................................... 6

FIGURE 1 - COMPARISON OF GRID BOUNDARIES AND CUTS ..................................................... 16

Tables

TABLE 1 — COMPONENTS OF THE EVOLVING PREFERRED OPTION .......................................... 9

TABLE 2 — SYSTEM OPERATOR AND GRID DEVELOPMENT VOLTAGE STABILITY LIMIT

ASSUMPTIONS .......................................................................................................... 15

Glossary

BHL-WKM The Brownhill to Whakamaru circuits, built to 400 KV standard but currently operated at 220 kV

CapexIM The Commerce Commission’s Transpower Capital Expenditure Input Methodology Determination

GSC Grid Support Contract, used for non-transmission solutions

GZ Grid zone

MCP Major Capex Proposal as defined in the CapexIM

Mvar Mega volt ampere reactive, a unit of reactive power

N-1, N-G-1 Security standards – described in footnote 6

Rankine A type of coal/gas plant owned and operated by Genesis Energy at Huntly

SVC Static var compensator

Transpower Transpower New Zealand Limited, owner and operator of New Zealand’s high-voltage electricity network (the national grid), and system operator

UNI Upper North Island

WUNI Waikato and Upper North Island

WUNIVM Waikato and Upper North Island Voltage Management

1 Introduction and background

Need for the WUNIVM Investigation

There have recently been actual and announced decommissionings of major generation

plant in the Upper North Island, being the region north of Huntly and including Auckland.

These so-called ‘thermal decommissionings’ are:

• 380 MW Otahuhu combined cycle unit (ceased generation in September 2015)

• 175 MW Southdown generation station (ceased generation in December 2015)

• 500 MW Huntly ‘Rankine’ units 1 and 2 (announced to be retired end of 2022)2.

These are significant changes for the New Zealand power system. Transpower conducted a

number of analyses in late 2015 and early 20163, and identified issues with both voltage

management in the Waikato and the Upper North Island, and thermal transfer into the

region4.

The voltage management issues are being investigated as the Waikato and Upper North

Island Voltage Management (WUNIVM) Investigation. This investigation is progressing

under the Commerce Commission’s Major Capex Proposal (MCP) process. We undertook

our long-list consultation under that process in July 20165 and the short-list consultation is

scheduled for the first half of 2019. While we have not yet fully developed our preferred

option to address the WUNIVM need, we have made significant progress.

The principal issue is voltage stability in the Upper North Island, making the investigation

complex and quite technical in nature. Transpower has analysed the issue against:

• the ‘N-G-1’ security standard, which requires the system to be robust to a single

credible contingency (asset failure) while one generator is out of service

• both static and dynamic voltage stability limits

• both expected (‘P50’) and prudent (‘P90’) peak demand forecasts.

2 Genesis Energy have recently made announcements about its plan to phase out coal use, but not yet any commitments on Rankine retention post 2022.

3 See www.transpower.co.nz/upper-north-island-generation-decommissioning-report-and-appendices and www.systemoperator.co.nz/activites/current-projects/impact-thermal-generator-decommissioning.

4 Investigation has determined that the evolving preferred option to voltage management would also have thermal transfer benefits, that would push any remaining thermal need out to the late 2020s or beyond. The term ‘thermal’ is used here to describe transmission limits to MW transfer, called thermal because the limit is related to the temperature of the conductors. Not to be confused the term ‘thermal decommissioning’, called thermal because it related to gas and coal fired generation plant.

5 Available at www.transpower.co.nz/waikato-and-upper-north-island-voltage-management-investigation. See also the WUNIVM Summary of and Response to Submissions document of October 2016 at www.transpower.co.nz/waikato-and-upper-north-island-voltage-management-submissions.

The summary results are illustrated in the following figure:

Figure 1 - WUNIVM need graph

In the figure, the red and green lines represent different levels of system security6, and the

two blue lines represent different load growth forecasts. As these are voltage stability limits,

the consequence of exceeding them could be severe, with the strong possibility of

widespread voltage collapse of the grid, resulting in extensive black-out and the need to

systematically ‘black start’ the grid.

The WUNIVM investigation has refined its analysis of the voltage stability limits that underpin

the need, which are presented for information in Appendix A.

The system operator has a principal performance obligation to avoid the cascade failure of

assets, and if required would implement a stability constraint on the market to prevent a

dispatch that could lead to cascade failure7. However, without investment in generation,

transmission or non-transmission solutions to raise the stability limit as demand grows and

6 The ‘N-1’ security standard requires the system to be robust to a single credible contingency (asset failure), and the ‘N-G-1’ security standard requires the system to be robust to a single credible contingency while one generator is out of service.

7 Such a constraint already exists as ‘UPPER_NORTH_ISLAND_STABILITY_P_1D’. For information see the latest manual constraint file at www.transpower.co.nz/system-operator/operational-information/security-constraints.

2500

2600

2700

2800

2900

3000

3100

3200

3300

3400

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

WU

NI d

em

and

an

d lo

ad li

mit

sA

nn

ua

l pea

k M

W (

in w

inte

r)

N-G-1

N-1

2 Huntly Rankine units in service Retired

generation plant is retired that constraint would need to be tightened to the point where it

could severely constrain the market.

Our analysis indicates that the power system would not be able to supply the peak Upper

North Island load from around winter 2025, even with two Rankine units available to provide

active and reactive power support, under the N-G-1 security standard and the prudent

forecast we use for planning purposes. The need grows from then at about 50 MW

equivalent per annum.

In addition, once the Rankine units retire, there is a much larger quantum of need, in the

order of 400 MW equivalent and increasing against the prudent forecast we use for planning

purposes. Even under the expected forecast, the quantum of this need is over 100 MW

equivalent and increasing. Given that the Huntly Rankine units may be decommissioned at

the end of 2022, this need has to be met before winter 2023. Given the time for analysis,

regulatory approval, procurement, build and commissioning, these need dates are tight.

Impact of Genesis Energy’s decisions on Huntly

This supply and demand situation has not materially changed since April 2016 when

Genesis Energy announced its plans to decommission the Huntly units 1 and 2 (the Rankine

units) at the end of 2022. There have been no firm commitments to new generation or to the

deferral or retirement of existing generation at, or north of, Huntly.

Genesis Energy has recently announced that it will stop using coal to generate electricity

except in exceptional circumstances by 2025, and it will stop using coal entirely by 2030.

This is not a commitment to retaining the Rankines past 2022, but does imply that they could

extend their life as coal plant until 2025, and as gas plant for normal market operations

beyond that.

It is possible that Genesis Energy will make a firm commitment to the Rankine units’

retirement, extension or replacement during the course of the WUNIVM investigation. If, and

as, that occurs, we will need to reflect that in the investigation. A possible outcome of a

commitment to Rankine unit extension or replacement would be deferral of WUNIVM

investment timings. In the meantime, we plan to proceed on the basis of the Rankines being

retired in 2022, as no commitment has been made otherwise.

We are also aware that if the Rankine units remain and we did not invest, then the existing

voltage stability constraint into Auckland would lead to Huntly units being constrained on.

This would initially be only for winter peaks but, as demand grows, for increasing proportions

of the year. Operationally, this could create market power issues. In investment terms, it

could ‘lock in’ a fossil fuel future, as there are few options for new large renewable

generation north of Huntly.

Evolving view of the preferred WUNIVM investment option

Under the Commerce Commission’s Transpower Capital Expenditure Input Methodology

Determination (CapexIM)8, we consider investment options and select a preferred option

from amongst them, justified by the CapexIM’s investment test. The preferred option is that

for which funding approval is sought. Each option is considered in the context of a longer-

term grid development path that contains modelled projects, being future potential

investments that are not part of the investment option but could affect the options and the

choice of preferred option.

This update provides information on the WUNIVM investigation’s evolving preferred option,

in Section 2. The modelled transmission projects that provide a context for this are

described in Section 3. We conclude with an explanation of our next steps in Section 4.

This update does not provide the reliability, technical or economic rationale for the evolving

preferred option, nor alternative options considered. That will be explained in detail in our

short-list consultation on the WUNIVM preferred option in the first half of 2019, which will

include the results of continued investigation plus the impact of updated demand forecasts

and any new generation commitments announced by then. Under the Capex IM, we are

required to justify investments with reference to the Grid Reliability Standards and based on

economic cost benefit analysis, which could include consideration of competition benefits9.

We provide this update in the interests of transparency and market information: readers

should consider it as a working draft as we test and develop our proposals, cost them more

fully, and continue our technical, reliability and economic analysis.

Please note that this is an information not a consultation document. Any requests for further

information should be directed to Conrad Edwards, WUNIVM Investigation lead,

Transpower, preferably by email to conrad.edwards@transpower.co.nz.

8 Available at http://www.comcom.govt.nz/regulated-industries/input-methodologies-2/transpower-input-methodologies/.

9 Clause D5(1)(k)

2 WUNIVM’s evolving preferred option

The WUNIVM investigation is proceeding and plans to put our preferred option to formal

short-list consultation in the first half of 2019, as part of the MCP process. The evolving

preferred option outlined below may change:

• as this WUNIVM technical, reliability and economic analysis proceeds

• if any generation commitments are made at Huntly or north

• with any material revision to demand forecasts

• consequent to submissions received on our short-list consultation.

The WUNIVM long-list consultation noted that the need is such that it is unlikely to be met by

a single solution – a range of components, commissioned at various locations and over time,

is likely to be required. The investigation has confirmed this, with the evolving preferred

option being for a range of components as shown below.

The components of our evolving preferred option are listed in Table 1:

Table 1 — Components of the evolving preferred option

Components of the evolving preferred option Expected need date

Shunt capacitors in the Waikato 2020

Series capacitors on the BHL-WKM circuits 2022

Demand management scheme to manage high impact, extremely low probability contingencies

2022

Additional components to manage voltage stability 2024+

These components are described below. Longer-term investment requirements are

discussed in section 3.

Shunt capacitors

Our evolving preferred option includes installation of 250 Mvar of shunt capacitors in the

Waikato.

At the pre-contingency load limit, the Waikato region already experiences low voltages,

visible as a dip when the voltage profile across the North Island is plotted. When the last two

remaining Huntly Rankine units retire, the voltage dip significantly reduces the voltage

stability limit in the WUNI region, impacting through-transfer to the Upper North Island (UNI).

In order to support higher peak active power (MW) transfers into the UNI region, the voltage

plane across the Waikato needs to be increased. An increase to the voltage plane would

also alleviate the reactive power losses on transmission circuits by reducing the amount of

reactive power transfer from distant generation or reactive support devices.

Shunt capacitors are the cheapest means of providing the reactive support necessary to

flatten the voltage plane.

250 Mvar of shunt reactive support is needed prior to the planned Rankine retirement in

2022, and would be needed even if the Rankine units remain in service from around 2022 in

a P90 prudent demand forecast. Given the long lead time to implement voltage support

infrastructure, and the uncertainty of both load growth and the Rankine units’

decommissioning date, we would plan to build them as early as we can, which is 202010.

The shunt capacitors would be installed in existing Transpower substations in the Waikato.

Other shunt capacitors may be required later, as described in section 2.4, on additional

components to manage over- and under-voltage.

Series capacitors

Our evolving preferred option includes installation of a series capacitor with 45%

compensation on each of the 220 kV Brownhill to Whakamaru (BHL-WKM) overhead

circuits. These are the 400 kV capable circuits built as part of the North Island Grid Upgrade

project.

Series capacitors would reduce the impedance (electrical length) of the line to improve

dynamic voltage stability in the Upper North Island region by:

• improving load division among the many parallel lines from the south into Auckland

• diverting power flow from heavily loaded circuits to the Brownhill to Whakamaru

circuits

• raising the WUNI thermal transfer limit

• improving dynamic system response

• reducing transmission losses during high transfers.

The series capacitors both improve dynamic voltage stability and allow higher MW transfer

into the UNI region without overloading the lower capacity parallel circuits. We own a

suitable site under the BHL-WKM line.11

The series capacitors are needed by the first winter following the Rankine retirement,

expected to be winter 2023. We would build them faster, if possible, to mitigate the risk of

an early generation plant retirement or transmission build delays.

10 To expedite this, and because these shunt capacitors are required even if the Rankine units are not retired, and are under the $20 million threshold, we plan to fund all the shunt capacitors through base capex. Within the WUNIVM MCP therefore we expect that these shunt capacitors will be excluded from the preferred option but included as modelled projects.

11 We maintained this site for the purpose of installing series capacitors, which were envisaged in the development plan of the 2009 North Island Grid Upgrade, with a then-estimated need date of 2021.

The reduction in transmission losses that series capacitors provide is such that it may be

economic to install additional capacitor banks to capture more loss benefits, dependent on

the trade-off with equipment cost.

Demand-management scheme

Our evolving preferred option includes consideration of the implementation of a post-

contingent demand-management scheme to cover specific extremely low probability

contingencies.

The series capacitors, as explained above, would divert more power through the high-

capacity Pakuranga to Whakamaru circuits, consisting of the Pakuranga to Brownhill

underground cables and the Brownhill to Whakamaru (BHL-WKM) overhead lines. These

circuits are built to a very high reliability standard, and have not had a fault in over five years

of operation12. The worst case ‘N-G-1’ event after the Huntly Rankine units have retired

would be if all the following events occur simultaneously:

• very high demand, such as on an extreme winter evening peak

• a tripping of one of the BHL-WKM circuits

• Huntly Unit 513 is out of service14

If this unlikely combination did happen it could – unless guarded against – cause voltage

collapse.

We could install primary transmission reactive plant specifically to safeguard against voltage

collapse if this unlikely combination of events occurred, but it would be very expensive to do

so. Our investigation has identified a lower-cost smart grid option in the form of a rapid post-

contingent demand management scheme. One option would be to arm this scheme (i.e. put

it into operation) only if there is a Huntly Unit 5 outage AND load in the UNI region is

exceptionally high15 (and maybe other similarly unlikely combinations of events). In the

unlikely event of a trip of either of the highly reliable Pakuranga to Whakamaru circuits while

the scheme was armed, sufficient UNI load would be shed almost instantly to avoid

widespread voltage collapse. That load would then be reinstated once the system operator

had stabilised the system.

While we understand the importance of maintaining a highly reliable electricity network we

also recognise the need to do this in an efficient manner to not unduly impose costs on

consumers. In this case, our preliminary view is that the risk of the combination of events

outlined above is sufficiently small to not warrant large-scale capital investment. The low

12 We have on occasion deliberately taken circuits out of service overnight to manage over-voltage issues. 13 Huntly Unit 5 is Genesis Energy’s gas-powered 400 MW generation plant. 14 We assume for planning purposes that at times of high load Huntly Unit 5 would be made available to the

market, unless it is under maintenance. 15 Our modelling indicates that it would never need to be armed under P50 expected load growth scenarios. It

would thus in effect be a safeguard against higher than expected load growth.

likelihood the scheme would ever need to be armed, and if it was, the very low likelihood it

would ever need to operate, is expected to make such an approach highly economic

compared to safeguarding the risk with much more expensive dynamic reactive plant. The

cost of this additional dynamic plant is estimated to exceed $50 million (2018 dollars).

That said, we also acknowledge that there are risks and technical challenges that we need

to investigate further to assure ourselves of using this approach. We will continue to explore

the merits of such a demand management scheme, and will justify it – or not – in our short

list consultation paper in the first half of 2019.

Additional components to manage voltage stability

The three components listed prior – shunt capacitors, series capacitors, and a demand

management scheme – would address the voltage need immediately after Rankine

retirement in 2022, i.e. winter 2023. Beyond that, without generation investment, we would

need additional voltage management as demand grows towards the new voltage stability

limit. Needs would be greater without a demand management scheme to cover specific

extremely low probability contingencies.

Our evolving preferred option anticipates some operational control will be required to

manage such issues, along with additional reactive support components. These

components are likely to include some more shunt capacitors in the Waikato and or one or

more dynamic reactive plant in the Waikato and Upper North Island before 2030. This could

include synchronous condensers, static var compensators (SVCs), STATCOMs or thyristor-

controlled reactors, or other types of reactive plant. The WUNIVM investigation is continuing

to explore the complex issues around these needs.

We will provide an update on this as we investigate further what additional components are

required to manage voltage stability post winter 2023. Depending on the outcome of this

continuing investigation, there may be further opportunities for grid support contracts16 for

non-transmission solutions to provide services post 2023.

In July 2017 we announced in our customer newsletter that “We had previously planned for

a non-transmission procurement process to start in 2017, on the basis that such a solution

could defer the major investments beyond the need date of winter 2023. However, the

emerging solution has deferred this need by three to four years. We now expect to start the

non-transmission solutions procurement process in 2020 or later (subject to Commerce

Commission approval), and have adjusted our timetable accordingly”. This remains the

case, and could include voltage support as well as transmission deferral GSCs.

16 Details on Transpower’s Grid Support Contract (GSC) product are available at www.transpower.co.nz/grid-support-contracts.

3 Modelled projects

As described in Section 1, under the CapexIM each investment option is considered in the

context of a longer-term grid development path comprised of ‘modelled projects’. The

modelled projects are new assets or changes to existing assets, that are not part of the

investment option proposed, but could affect the options and the choice of the preferred

option.

Investigation has determined the evolving preferred option to address voltage management

needs outlined above would have thermal transfer benefits. With further demand growth we

will eventually reach thermal transmission capacity limits for transfer into Auckland17. In the

absence of significant generation commitments at Huntly or north, this is expected to occur

in the late 2020s under our P90 prudent load growth forecast, or in the 2030s under our P50

expected load growth forecast. Then we would need to augment the grid to increase

capacity between Whakamaru and Auckland.

There are many ways in which the grid could be evolved to increase thermal transfer limits

into Auckland, and these will be explored as modelled projects in our Short List Consultation.

We have investigated these options so far only to the point of reassurance that our emerging

preferred solution to WUNIVM is robust to different grid futures in the region.

Any future such transmission investment might be the subject of a future MCP proposal,

which could be expected to explore grid support contracts for transmission deferral.

It is possible that to support increased transfer, more reactive power plant may also be

required in this timeframe. If so, this too will be a modelled project in the MCP process

rather than part of the preferred option for which we will be seeking funding approval.

4 Next steps

The WUNIVM investigation is continuing to refine our preferred option, especially in the

areas of additional components to manage voltage stability. We are continuing with testing

and developing our proposals, costing them more fully, and continuing our technical,

reliability and economic analysis.

We are planning on pulling these together into a short-list consultation in the first half of

201918.

17 The first thermal transmission constraints are overloading of the two existing underground cables from Brownhill to Pakuranga, and overloading of the 220 kV circuits between Hamilton, Ohinewai and Whakamaru.

18 The shunt reactive capacitors will be included as base capex and so treated as modelled projects in our MCP submission, as noted above.

Voltage stability limits

This section provides an update on the voltage stability limits into the Waikato and Upper

North Island region (WUNI). These have been refined through the WUNIVM investigation

and may be of interest to market participants.

WUNIVM Voltage stability limits

This section describes the voltage stability limits for the combined Waikato and Upper North

Island (WUNI) region.

The voltage stability load limit19 in WUNI is around 3,115 MW under N-G-1 security criteria if

two Huntly Rankine units are available providing full active and reactive power.

After decommissioning of all Huntly Rankine units, the voltage stability load limits reduce to

around:

• 2,970 MW under N-1 security criteria

• 2,725 MW under N-G-1 security criteria.

The above numbers are approximate because they are sensitive to the assumptions

including but not limited to load profile at various grid exit points, reactive support available

from existing reactive plants and generation dispatch level.

The need graph in section 1 (Figure 1) shows the WUNI voltage stability load limits plotted

along with the 2017 winter peak load and winter prudent peak demand forecast from 2017 to

2027. The worst case outage for:

• N-1 is the loss of a 220 kV Pakuranga–Whakamaru circuit.

• N-G-1 is the loss of a 220 kV Pakuranga–Whakamaru circuit when Huntly unit 5

is not available.

19 The load limits presented includes a 5% margin in accordance with standard international practice to account for the uncertainties in assumptions.

Transpower’s operational and grid planning voltage stability limit assumptions

Transpower’s operational and grid planning functions use different assumptions in

calculating the voltage stability limits that each publishes:

Table 2 — Operational and grid planning voltage stability limit assumptions

Assumptions Operations Grid planning

Purpose Real-time operation Long-term planning

System conditions Actual as known As expected year ahead

Assets in service Actual plant, allowing for outage plans

Actual plus committed plant (and less expected decommissionings)

Assume all but the largest generation and reactive plant in service (N-G-1)

Reactive load Measured Mvar Modelled motor load20

For these reasons, the voltage stability limits expressed by the system operator and

Transpower’s grid planning documents will typically differ.

Comparison of grid boundaries and cuts

The following schematic map illustrates – around the WUNIVM region – the differences

between the:

• System operator’s grid zone (GZ) boundaries

• Grid planning regional boundaries (WUNIVM encompasses the Waikato and

Upper North Island regions)

• Cut of the current Upper North Island Stability Constraint21.

20 Based, in the case of WUNI, on recently commissioned motor load surveys. 21 See www.transpower.co.nz/system-operator/operational-information/security-constraints

Figure 2 - Comparison of system boundaries and constraint cuts

As illustrated here, the WUNI demand and load limits expressed above (and in the y-axis of

Figure 1) are equivalent to GZ1 + GZ2 + GZ3 – KIN (Kinleith) load, and the Upper North

Island Stability Constraint is based on power transfer into the GZ1 + GZ2 region.