1/30 Rate of Change of Frequency (RoCoF) Modification to the Grid Code DOCUMENT TYPE: Decision Paper REFERENCE: CER/14/081 DATE PUBLISHED: 4 th April 2014 QUERIES TO: [email protected]The Commission for Energy Regulation, The Exchange, Belgard Square North, Tallaght, Dublin 24. www.cer.ie
30
Embed
Rate of Change of Frequency (RoCoF) Modification to the ... · 6/30 2 Overview 2.1 DS3 & RoCoF EirGrid has proposed the RoCoF modification in order to facilitate the delivery of the
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1/30
Rate of Change of Frequency (RoCoF) Modification to the
for them to incur particularly as the benefits of the RoCoF change will accrue to wind
plant only. In addition, generators have outlined concerns with the definition of the
modification particularly in relation to the timeframes for recording compliance, while
some generators have also challenged the interpretation of the results of the desktop
analysis carried out by DNV KEMA for EirGrid.
The vast majority of wind generators have no difficulty with the technical standard
proposed and have an obvious commercial benefit to its implementation. They argue
that the longer the delay (and the greater the impact on curtailment) the less likely it is
for developers to be able to make investments. There are also issues associated with
the distribution system which must be addressed during the implementation phase of
any change to RoCoF requirements.
2.2 System Operation
Changes to Grid Code RoCoF requirements are required in order to increase the SNSP
to the eventual 75%. Without RoCoF changes it is unclear what the final SNSP could be.
Significantly credible or detailed alternatives to increasing SNSP through RoCoF
changes have not been put forward to date by either the TSOs or members of the
Working Group although it is apparent that mechanisms which increase system inertia
without impacting on curtailment levels could play a role in increasing SNSP (e.g. lower
minimum generation levels of generators). However it is not clear that such mechanisms
are either more straightforward, less costly or would have as great an impact as a
change to the RoCoF standard in the Grid Code. Given the interdependencies in the
system improvements proposed by the DS3 programme there is a limit to what can be
delivered through System Services, other Grid Code modifications and TSO operational
improvements. For the new RoCoF standard to have any impact, the TSOs have stated
that all (or almost all) generators must be able to comply with the new standard. If the
system experiences a RoCoF of 1Hz/s and one generator tripped, the level of system
RoCoF would increase possibly resulting in another generator (who withstood the
original RoCoF event) tripping, increasing the RoCoF yet further, and so on. This
cascade effect would threaten system security.
Therefore it is essential that all generators are compliant with the standard if the system
is to operate at a higher SNSP (above the 50% current operational limit). At such an
SNSP, any generators that were not compliant would have to be taken off the system
during high wind events to mitigate the risk of such a system failure and/or wind would
have to be curtailed.
Proving compliance with a higher RoCoF standard than that currently required in the
Grid Code will be an issue for the TSO. There is no test that can reliably check that a
generator will withstand a high RoCoF event. Therefore the TSO is almost entirely
reliant on the generator’s assessment and “certification” of the unit’s capability. It is for
this reason that the generator studies (discussed below) must be robust and of sufficient
8/30
quality to convince the TSO that it is safe to operate the system with a high SNSP when
that unit is on the system.
2.3 Conventional Generators
Conventional generators have two main concerns aside from the ability to remain
synchronised. These are that a high RoCoF event will cause a catastrophic failure of a
unit – which is primarily a safety concern for station staff – and that repeated high
RoCoF events will negatively impact the commercial life of the plant. The risk of
catastrophic failure is considered by the CER’s consultants to be “highly unlikely” on the
basis that units can be expected to undergo more severe network fault events without
such catastrophic failure. The impact on the commercial life of the plant is highly
dependent on the frequency of high RoCoF events. If they are infrequent, as is
expected, then there will be minimal impact on the life of the unit. The frequency of such
events cannot be forecasted with accuracy. That said, there is a greater likelihood of
increased wear and tear on conventional plant in the coming years as levels of wind on
the system increase, even in the absence of RoCoF events. This will be due to
increased cycling of plant to accommodate wind and other priority-dispatch units. While
on one hand it can be argued that increasing the grid code RoCoF requirement will drive
further wear and tear, it would not be possible to separate out RoCoF associated wear
and tear from normal, non-RoCoF wear and tear. In any event, conventional generators
must take note of the clear direction of European and National policy which favours
incremental movement towards a low carbon electricity system as well as the explicit
requirements of the Renewables Directive (2009/28/EC) to minimise curtailment of wind.
By implementing these policies through binding national targets, the policy makers have
made clear the importance of achieving these objectives regardless of the impact on pre-
existing investments. This policy will result in a system with a high penetration of wind
and a system which has different technical requirements to those currently prevailing.
Indeed failure to achieve the target will not only result in a higher carbon electricity
system, but also significant fines for Ireland from non-achievement of its binding targets.
It is estimated that these fines could be in the significant.5 The purpose of the Grid Code
is to establish the minimum standards required to maintain system security and the
electricity system in 2020, and the years leading up to it, will require a generation fleet
capable of withstanding higher RoCoF events than the current standard.
In order to ascertain generators’ ability to withstand a RoCoF event of 1Hz/s over 500ms
studies will need to be carried out on every unit on the system. The only alternative at
5 The Renewables Directive provides a formula for Member States to calculate their trajectory to meeting
their target in 2020. Ireland is currently on target with regard to its trajectory. Failure by Member States to meet their individual binding target by 2020 (in Ireland’s case 16% of total energy met by renewables) would result in the European Commission taking infringement proceedings. While the scale of such penalties cannot be known at this point, the Sustainable Energy Authority of Ireland (SEAI) estimates that non-compliance costs could amount to around €140 million to €210 million for each 1 percentage shortfall relative to Ireland’s EU target of 16% of energy from renewable sources by 2020.
9/30
present is to wait for a RoCoF event(s) to take place, with the associated risk of system
outages or system failure. The main difficulty however with the study approach is that
such studies have not been carried out previously and it is therefore not possible to
utilise prior international experience. The costs which have been indicated to the CER
while high could be considered to be reasonable were a large scope of specialised work
required, albeit that it is accepted by CER that out of necessity, Ireland will be a “first-
mover” internationally in carrying out such studies, with the associated advantages and
disadvantages of first mover status.
It is noted that several generation units on the island have similar equipment and there
may be some scope for co-ordinating the studies of different units (i.e. plants with the
same type of turbine) and the required network modelling but it appears that there will be
a significant element of each study that will be unique to the unit in question. The studies
will only determine if a unit can withstand a 1Hz/s RoCoF event. Therefore it is possible
that the studies may conclude that it is not possible to increase the standard or that an
implementation phase is required (retrofitting of plant etc.). It is also noted that some of
the manufacturers active in Ireland have also built units for the Danish market. Denmark
has a standard of 2.5Hz/s for new units (and previously had a standard of 2.0Hz/s) –
manufacturers may be able to make some use of the studies undertaken in the design of
these units.
2.4 Wind Generators
Wind farms, with some exceptions, do not have any issue with the proposed standard
and the manufacturers have confirmed they could comply with the proposed RoCoF
standard.
The main issue for wind is that the longer the delay in resolving this issue the higher
wind’s curtailment levels are likely to be. Therefore there is considerable concern in the
industry that a delay to the implementation of the proposed RoCoF standard will have a
significant impact on the commercial viability of wind projects.
2.5 Distribution System
The DSO currently uses RoCoF protection to prevent islanding on the distribution
system. Islanding occurs when a section of the distribution system becomes separated
from the rest of the system but remains live. A RoCoF of 1Hz/s is not compatible with
current DSO practice. If no changes were made on the distribution system a high RoCoF
event could result in large parts of the distribution system tripping creating a cascade
effect across the entire transmission network. The DSO supported the TSO’s
modification and based on their analysis consider that the current protection settings can
be modified to allow for the 1Hz/s standard. There is a considerable amount of further
work to be undertaken in terms of implementation and engagement between TSO and
DSO to work out remaining technical details. While the amount of outstanding work
10/30
should not be underestimated there does not appear to be any issues which would
prevent the implementation of the TSOs’ proposed RoCoF standard.
11/30
3 CER Decision
3.1 Summary
The CER approves the modification in principle, but it will only come into effect following
confirmation from EirGrid that, from a system security perspective, it can be
implemented. To determine this there will be an industry implementation project made up
of three strands; TSO & DSO implementation; Alternative solutions; and generator
studies. The implementation of RoCoF will be phased over a period of 18 to 36 months,
with higher priority units being required to complete their studies first. The overall
industry project will be coordinated by an independent consultant and overseen by the
CER. Incentives will be implemented and will be progressed through the established
SEM process.
3.2 Approval of Modification
Having reviewed EirGrid’s recommendation, submissions from industry, the independent
report from PPA, the responses to CER/13/143, further submissions made by individual
generators and the EAI and the national and European policy background, the CER
considers that it is critically important to increase the RoCoF standard in the Grid Code
in order to facilitate the achievement of Ireland’s legally binding renewables targets.
Delivery of higher levels of wind penetration on the electricity system is a clear national
policy and it is important that the CER does not obstruct this policy. However the CER is
cognisant of the fact that delivery of policy should only take place without impacting on
the quality, reliability or safety of electricity supplies. Therefore it is important that in
approving the RoCoF modification in principle that the modification can be implemented
without a significant risk to the safety and reliability of the electricity system. As outlined
in the PPA report, published alongside the consultation paper, there is a level of
uncertainty regarding the technical capability of the Irish generation fleet and the
potential requirements to undertake work to comply with the higher RoCoF standard, the
CER therefore considers it prudent to propose to delay implementation of the RoCoF
modification (MPID 229) for a period of time to allow the required studies and works to
be carried out.
Therefore, the CER approves the modification, as proposed, in principle but the CER will
not give effect to the new standard in the Grid Code until it has received confirmation
from EirGrid that, in its professional judgement, a sufficient number of generators can
comply with the standard to allow EirGrid to safely operate the system in a manner
reliant on the new RoCoF standard. Generators shall be required to make a declaration
to EirGrid regarding their level of compliance within 18 to 36 months of the publication of
this paper.
12/30
3.3 Phased Implementation
The deadline for the declaration of compliance (or submission of a derogation) will be
phased according to the importance, in terms of system operation, of each unit. The
CER directs EirGrid to categorise each unit into i) high-priority ii) mid-priority iii) low-
priority and iv) exempted units. The TSO will use its judgment in making these
categorisations but will have regard to the following criteria:
1. High-priority: relatively high run-hours; frequently constrained on; frequently run
at times of high wind
2. Mid-priority: those units not falling into the other categories
3. Low-priority: low run-hours; infrequently constrained on; rarely running at times of
wind generation
4. Exempted units: units which are soon to retire; very low run-hours; infrequently
constrained on; very rarely running at times of wind generation, units which
EirGrid’s operational experience shows would have historically experienced and
ridden through high RoCoF events.
5. New units: new units will be required to declare compliance during the
commissioning process
The formal commencement of the RoCoF implementation project will be notified to
industry through the CER’s website. High-priority units must make a declaration of
compliance (or submit a derogation) within 18 months of this commencement of the
RoCoF implementation project. Low-priority units must make a declaration of compliance
(or submit a derogation) within 36 months. Mid-priority units must make a declaration of
compliance (or submit a derogation) within 24 months. Exempted units will not have to
complete a study though they may choose to do so, if they wish. For the avoidance of
doubt any unit may apply for a derogation; however a necessary component of such an
application will be a study detailing the nature of the non-compliance and the remedial
actions taken to attempt to rectify the non-compliance. Exempted units that cannot
comply with the requirements will be required to formally make an application for a
derogation, though they will not be required to enclose a RoCoF study with their
application. It is not envisaged that any derogations will be assessed by the CER prior
to the 18 month deadline.
The TSO shall categorise each unit based on the criteria set out above and seek
comments from generators prior to submitting their recommendation to the CER for
approval. For the avoidance of doubt any unit categorised as “exempted” by the TSO,
and approved as such by the CER, is so categorised without prejudice to the CER’s
decision in relation to the derogation application the unit ultimately submits.
13/30
The TSO will also assess the possibility of operating the system at a higher penetration
of non-synchronous generation where a portion of the generation fleet has demonstrated
compliance with the new standard and a portion has not. The TSO will assess the
viability of relying on the dispatch of compliant generators at these times although it is
acknowledged that there may be implementation issues to be overcome. Under such
circumstances non-compliant generators would be considered technically unavailable
when the SNSP (or equivalent metric) is over 50%.
3.4 Generator Studies
As discussed above generators have indicated that they must undertake technical
studies in order to determine their unit’s ability to meet the proposed RoCoF standard. In
discussions with generators, individually and as part of the Working Group it is
understood that such studies should take between 12 to 18 months to complete.
However, of the conventional generators who responded to the CER/13/143 the general
view was that the proposed 18 month deadline for completion of studies was
unreasonably short. Generators cite difficulties such as the technical complexity of the
studies themselves, the requirement to rely on OEM’s6 (multiple OEMs in some cases)
actively engaging with the generator and resource constraints within the OEMs even if
active engagement is achieved. The CER has considered generators’ responses and
while it is accepted that such studies are significant and complex pieces of work, delays
in the completion of the studies will directly impact on the level of curtailment faced by
wind generators, their financeability, the national renewable targets, and delay savings
to consumers associated with an increased SNSP. Therefore the CER wishes to
proceed as quickly as possible while acknowledging the difficulties faced by
conventional generators.
Process for carrying out generator studies
As set out above EirGrid will identify the relative priority of each unit. EirGrid will
engage with industry on these priority classifications and will submit them to the
CER for approval.
In order to ensure consistent delivery of studies and a fair and transparent
process, the overall industry project will be coordinated by an independent
consultant and overseen by the CER, possibly in co-ordination with UR. The
Independent consultant, in consultation with generators and the TSO, will set out
the requirements from EirGrid in terms of information provision to generators
(e.g. system scenarios) and the requirements from generators in terms of
carrying out the studies and the minimum level of information which the generator
will need to obtain from the OEM.
6 Original Equipment Manufacturer
14/30
The generator studies will be project managed by the generator concerned with
an agreed report structure to the independent consultant to ensure consistency
across all studies. Specifically the TSO will input into the generator’s study at the
outset, at pre-agreed interim milestones and at the study’s conclusion.
It is envisaged that there would be a tripartite meeting at the outset of every
generator’s study with the TSO, generator and the independent consultant, but
the exact organisational structure will be determined by the Independent
consultant in discussions with the CER. The CER may attend these meetings as
it considers it appropriate. At this initial tripartite meeting the scope of the study
will be discussed based on what results the TSO considers necessary and on
what results the generator can reasonably deliver through the study. Interim
milestones will also be agreed at this initial meeting where progress will be
reviewed and any changes to the plan that may be required based on preliminary
results will be agreed.
Upon the conclusion of the generator’s study and submission to the Independent
consultant (who will review for consistency), the TSO will review it with a view to
determining the overall security of the system with the new RoCoF standard.
The independent consultant, in consultation with EirGrid and industry, will propose the
scope of the appropriate TSO network modelling at the outset of the project. EirGrid and
generators will assist by providing all reasonably required information. If required, the
CER, working with UR, will co-ordinate the sharing of information if required for reasons
of commercial confidentiality, ensuring that confidential information is protected at all
times. The objective of this modelling is to provide an agreed set of system scenarios
upon which the generator studies can be based and additionally to facilitate generators
appropriately scoping their projects.
Generators will be required to submit detailed project plans to the CER, the independent
consultant and give frequent updates to the independent consultant co-ordinating the
industry project. Generators will also be required to regularly publish public reports on
their progress – the publication of reports will be co-ordinated by the consultant. The
frequency of both the project updates and the public reports will be determined when the
project plans are being agreed.
3.5 TSO-DSO Implementation Project
As discussed above there is a significant programme of work required to implement the
new RoCoF standard on the distribution system. The DSO is directed to set out a project
plan and will provide regular updates to CER and the industry at appropriate DS3 fora,
the distribution code panel and the Grid Code review panel.
As part of the project governance of the overall RoCoF implementation project the TSO
and DSO will engage regularly to agree and monitor the delivery of the project. The SOs
15/30
will provide public quarterly updates on this work stream through the existing DS3
communications process.
The potential impact of higher RoCoF events on demand customers was raised in
bilateral discussions, noted in the PPA report, and by one of the respondents to
CER/13/143. From subsequent discussions with the system operators the CER
understands that it is unlikely that demand customers should be negatively impacted by
the new Grid Code standard. However, the CER considers it prudent that this issue is
monitored to ensure that quality of supply is not negatively affected.
The CER directs the DSO and TSO to monitor the impact of the new RoCoF standard on
demand customers and the quality of supply as part of the TSO-DSO Implementation
Project, and report any concerns, if any, to the CER.
3.6 Alternative Solutions Project
In light comments received during the extensive consultation with industry on this topic
the CER is of the view that there is a risk that lead time for implementation of a new
RoCoF standard could take longer than the 18-36 month deadline set out by the CER in
this paper. The CER must also be cognisant of the possibility that a new RoCoF
standard will not be implementable in a timely enough manner to affect the 2020 targets.
Therefore it is considered prudent that the CER direct EirGrid to investigate, and where
relevant propose and implement, alternative solutions to the inertia problem that RoCoF
would resolve. If complementary and realistic alternatives to RoCoF consistent with the
objectives of the DS3 programme can be delivered, they may mitigate any potential
impact of a delay to the completion of the RoCoF implementation project and may
ensure a somewhat higher SNSP than otherwise possible. The TSO will provide regular
updates on its progress through the existing reporting arrangements for DS3. The TSO
should engage with industry in developing the scope for this project and throughout. The
report will be submitted to the CER within 18 months of the publication of this paper. It is
envisaged that where the TSO can make recommendations or implement measures
earlier than 18 months, this should be done.
The TSO should consider the following areas as part of this project:
Measures to increase inertia on the system through network investments, storage
and strategic generation investment
Synthetic inertia
Changes to operational policy
Any other matters the TSO considers relevant.
16/30
3.7 Financial Arrangements
Generators have requested that the costs associated with the technical studies be
recoverable. In CER/13/143 the CER set out two options, no cost recovery or costs
socialised across all generators. The CER stated that its preferred option was for no cost
recovery. The CER remains of this view and accordingly will not be providing for cost
recovery of the studies.
Several generators have made representations to the CER, both before the publication
of, and in response to, CER/13/143, requesting that the costs associated with the
studies be recoverable. It is noted that this would be a departure from industry practice
as compliance with the Grid Code is the responsibility of the generator as are any costs
required to achieve or maintain compliance. Electricity systems across the world are
demanding greater flexibility from generators in response to initiatives to diversify
supplies and increase renewable (often intermittent) penetration. Ireland is no different,
indeed the requirement for flexibility is more acute, and against this background, the
CER is of the view that it is reasonable to expect improved flexibility from all generators
on the system. In some cases, it is appropriate to reward this flexibility and the DS3
System Services work stream is currently considering appropriate levels of payment in
this area. However in other areas it is appropriate that improved flexibility is mandatory
giving the changing nature of the generation portfolio.
That said there is also an argument to suggest that the RoCoF modification is slightly
different in nature from other Grid Code modifications in that the costs for delivering
compliance rest chiefly with conventional generators without any associated benefit from
making the required investment. In fact, delivering compliance is likely to result in a
negative commercial impact on many conventional generators as it actively facilitates a
displacement of conventional generation by wind generation. Notwithstanding the
national policy requirements, the commercial benefit of the successful implementation of
this modification will go to wind farms, whereas the successful implementation will
disadvantage the commercial position of conventional generators as they will be more
frequently displaced by wind generation. This coupled with the costs of the studies does
not incentivise timely implementation of the modification.
While the CER can appreciate the views of generators who feel they are required to pay
for expensive studies to prove compliance with a modification which will, upon
implementation, actively disadvantage their plant, the CER has nonetheless decided not
to allow cost recovery. On balance the CER does not consider that the arguments in
favour of cost recovery are sufficient to warrant a change in policy regarding Grid Code
modifications. It is current policy, and a condition of the generator licence, that all costs
associated with Grid Code compliance are the responsibility of the generator.
Furthermore, the RoCoF modification has arisen as a direct result of European and
national policy which favours renewable penetration. The purpose of the Grid Code is to
set the minimum standards necessary to ensure the safe operation of the system given
the nature of that system. As a result of European and national policy the nature of the
17/30
Irish system will be different in 2020 to the historically traditional system with a portfolio
of predominately conventional generation. Generators must meet the minimum
standards required of a system with a high penetration of non-synchronous generation,
including a higher RoCoF standard than previously required. Generators in signing up to
the Grid Code carry the risk of the cost of implementing changes to the Grid Code; if the
generator decides not to comply with the Grid Code then there is a clear signal that that
generator should not be a connected party to the Irish electricity system. Accordingly the
CER will not provide for cost recovery and generators should pay in full for the costs of
their individual studies.
However, the CER acknowledges that in addition to the costs associated with the studies
there will be operational costs associated with higher RoCoF events. Such costs may not
be recoverable through energy bids. Accordingly the CER and the Utility Regulator in
Northern Ireland will recommend that the SEM Committee request the TSOs to consider
and propose the introduction of a remuneration mechanism which may include a new
Harmonised Ancillary Services (HAS) rate for RoCoF. It is envisaged that all generators
demonstrating compliance with the 1Hz/s standard would be eligible for a period of time.
While it is hoped that the introduction of a remuneration mechanism would incentivise
the timely completion of the generator studies it is considered that a Generator
Performance Incentive (GPI) should also be applied. The CER has revised the proposed
design of the GPI to lower the daily charge and to phase the introduction of the full
charge. Units will become eligible for the GPI according to the deadline associated with
their categorisation as discussed in Section 3.3. The CER and the Utility Regulator will
recommend to the SEM Committee that a harmonised GPI of the form set out below be
applied on an all-island basis.
18/30
Formula:
(a – b) x (€5,500) x (d) x (e) = c
Where a=RoCoF standard; b=Unit’s RoCoF level; d=scalar associated with size of unit;
e=scalar associated with the period of time from the commencement of the