Wind Power Transmission Workshop 2005

    Wind Energy Delivery and Reliability

                  March 30-31, 2005
            Toronto, Ontario, Canada

        Workshop Summary Report

          Prepared by:    Dr. Geza Joos
                          McGill University

          Prepared for:   Canadian Wind Energy Association

          Date:           March 31, 2005
This report provides a summary of a wind energy delivery and reliability workshop held March
30-31 2005 in Toronto Canada. The workshop was organised by the following:

       •   The American Wind Energy Association (AWEA)
       •   The Utility Wind Interest Group (UWIG)
       •   The Canadian Wind Energy Association (CanWEA)

The preparation of this summary report was funded by Natural Resources Canada through the
Technology and Innovation Program as part of the climate change plan for Canada.
                                   EXECUTIVE SUMMARY

Wind energy in Canada at this time produces a very small portion of the electricity supply.
Canada’s total installed capacity of 444 MW satisfies only 0.2 % of the nation’s energy
demand. However, with new projects totalling at least 2000 MW coming on-line in the near
future, and more planned, it is expected that wind energy will cover 3 % of the energy needs of
this country by the year 2012. CanWEA and AWEA therefore identified a need for a workshop
to cover the more important issues related to wind energy delivery and reliability.
The workshop, held March 30 and 31, 2005, dealt specifically with North American and
European experiences and studies in wind energy, both on existing and planned installations. It
also addressed the issues of wind variability, reliability and wind forecasting. Finally, it
presented issues related to wind power integration into the electric grid, as well as
interconnection standards and codes. More specifically, the workshop provided perspective on
the following issues:
-   Requirements for integrating wind energy as penetration levels increase,
-   Estimation, through forecasting, of wind energy value and its variability and reliability,
-   Methods of compensating variability by associating wind with hydroelectricity, given that
    hydroelectricity covers 60 % of the energy needs in Canada,
-   Evaluation of the potential impact of wind generation on the operation of the electric grid,
    particularly voltage profiles, and on the reliability of the grid,
-   Grid interconnection codes and requirements.
The issues addressed in this workshop can be categorized into two classes: integration issues
and interconnection issues.
Integration covers issues related to the interaction of wind with other loads and energy sources
connected to the grid, namely impact on electric grid operation, voltage profiles, and grid
reliability. Integration is primarily an issue of economics, and how wind may be managed on the
grid to the financial benefit of both the utility and customer. Interconnection considers the
connection strategies of wind power to the electric grid, and its interaction with the grid. In
contrast to integration, interconnection deals mostly with technical issues, such as voltage and
frequency ranges of operation, reactive power control, wind farm contributions to grid voltage
support under grid fault condition, and the control of power injected into the grid.

Workshop speakers presented the results of wind integration studies in Canada (Ontario,
Manitoba and Quebec), the USA (New York State and Minnesota), and Europe (Denmark). Key
points of these presentations are summarised below.
-   In Ontario, a 2004 study by AWS Truewind examined the potential impact of integration of
    2000 MW of wind at 17 sites spread across the province. It concluded that capacity factors,
    (the ratio of energy produced to installed capacity) varied by season from 47% in winter to
    19% in summer, providing a good correlation with load on a seasonal basis. Conversely,
    the study found that correlation on a daily basis was less favourable, but that wide
    geographic placement of the turbines greatly reduced overall variability of output. The study
    found that requirements for load following were increased by only a modest amount.
Workshop Summary - Wind Delivery and Reliability                                 March 30-31, 2005

-   Manitoba is investigating the possibility of integrating 1000 MW of wind and considers wind
    as a commercial export opportunity. The province has sufficient capacity for domestic
    supply until 2020, and is looking at wind as a complementary energy source to its hydro
    plants. By using wind effectively, the province may be able to sell larger blocks of high-
    value (peak period) electricity from its hydroelectric facilities to customers in the United
    States. Extensive studies are being carried out to determine the capacity value of wind, the
    impact of short-term variability, and wind-hydro integration. The load balancing and energy
    shaping capabilities of hydroelectric plants are also being examined. It is expected that
    existing constraints, such as water reservoir levels in the short and long term, and energy
    scheduling will impose limits on penetration.
-   Quebec is in the process of initiating studies to determine the impact of integrating at least
    1500 MW of wind energy over the next 10 years. Among the issues under investigation are
    the impact of short-term variability, the accuracy of wind forecasting, wind-hydro integration
    and the capacity value of wind.
-   In New York State, a study by General Electric examined the impact of the integration of
    3300 MW of wind (spread over 37 sites, representing 10 % of total capacity) on system
    reliability. Results showed that the increase in required regulation reserve would be
    minimal, as would be the other impacts of wind, namely in terms of load following, power
    flows and transmission line congestion. Power ramp rates of wind turbines are similarly not
    a cause of concern. Studies of voltage profiles reveal in addition that wind turbines reduce
    the requirements for voltage regulation on the grid to which they are connected.
-   In Minnesota, a study assessed the impact on grid reliability and ancillary services of the
    integration of 1500 MW. The study showed that; a) system regulation increased by only a
    small amount; b) the cost of load following was very low, and; c) that the major difficulty lay
    in providing accurate day ahead scheduling. Good wind forecasting is therefore needed for
    proper integration.
-   Denmark has one of the highest wind penetration rates in the world (16 % of generation). A
    representative of a large Danish utility reported on the issues associated with integration.
    These include load following, the impact of extreme weather conditions, the need for better
    forecasting, and the reliance on interconnections with neighbouring jurisdictions for load
    balancing. It was pointed out that not all installed generating units are capable of providing
    load balancing, as combined heat and power units must primarily provide heat, It is
    expected that higher penetration levels may require some redesign of the power grid, as the
    existing power grid was not designed to handle variable generation.
-   Great Britain plans on integrating large amounts of wind generation in the near future.
    Studies show that this can be done in the short term with a minimal amount of modification
    to the existing grid, and with no significant integration cost having to be added to the price
    of electricity. The speaker also stressed the need for proper wind forecasting to allow
    scheduling of load balancing.
The shaping and balancing services that can be provided by hydroelectric plants was also
described, with reference to the Bonneville Power Administration service. Pumped storage and
compressed air were mentioned as being capable of providing this service, along with
combined heat and power units. Other issues included the use of available transmission
capacity on existing corridors, in particular high voltage dc lines.

Workshop Summary - Wind Delivery and Reliability                                  March 30-31, 2005

Presenters and participants all agreed that wind energy can and will be integrated in significant
amounts into the electric grid. It is expected to form an integral part of the total energy portfolio
in Canada and elsewhere. The main conclusions of the presentations and the ensuing
discussions are the following:
-   Wind penetration levels of up to 10 % of capacity will have a minimal impact on grid
-   Storage, shaping and balancing are key issues for proper integration of wind energy.
    Hydroelectric provides excellent complementarity in this regard.
-   Electric market structures and rules will have to evolve to allow a proper integration of wind
    energy and to give it a suitable place in the energy portfolio. In the transition period,
    government incentives are required to make wind energy competitive.

Grid codes have evolved significantly in the last few years as transmission grid operators
demand that the wind farms behave more and more in a manner similar to conventional
generators. Wind-specific interconnection standards and codes are either in place or under
development in a number of jurisdictions. These interconnection standards address a number
of wind-specific issues, including requirements such as low voltage ride through (LVRT).
Extensive studies done for the Alberta Energy System Operator examined interconnection
requirements to determine an appropriate grid code for its transmission system. It considered
voltage and frequency operating ranges, as well as power and reactive power specifications.
Regarding low voltage ride through requirements, the study indicated that a characteristic curve
similar to the one developed in Germany (E.On) would be suitable for its system.
AWEA is proposing that a similar code be adopted through the Federal Energy Regulatory
Commission (FERC). It proposes reactive power control and power factor control ranges, under
steady state and transient conditions. Voltage regulation capability may have to be provided by
means of variable reactive power. It is expected that FERC will adopt the AWEA recommended
grid code later this year.
Wind turbine generator manufacturers discussed the methods they are proposing to meet grid
code requirements. It is generally agreed that grid codes can be satisfied either by means of
power electronics embedded in the turbine generator or external compensating systems,
depending upon the design of the turbine generator. Most manufacturers offer generators of
different types incorporating power electronics. Independent suppliers offer external systems,
comprised of capacitors, fixed or mechanically switched, and static var compensators. They
demonstrated how to configure these different types of systems to satisfy grid code
requirements, particularly as seen from the grid at the point of connection. Low voltage ride
through, reactive power and voltage regulation, and steady state and transient voltage
operating ranges issues are considered.
From presentations and discussions, the following conclusions are drawn:
-   Grid integration issues at the wind turbine generator level are well understood, including the
    required ranges of operation in terms of grid voltage and frequency, as well as the need for
    low voltage ride through capabilities. Manufacturers are willing to implement these
    requirements on their products, while insisting on the specific features of wind turbine

Workshop Summary - Wind Delivery and Reliability                                  March 30-31, 2005

    technologies that make them different from conventional synchronous machine based
    power generation.
-   There are no objections in principle to harmonizing the important aspects of grid codes.
    However, both utilities and manufacturers mention that uniform requirements for all aspects
    of interconnections and in all installations and jurisdictions may not be desirable or
-   It is generally accepted that wind farms need to and can participate in the voltage regulation
    of the electric grid at the point of connection to the grid. However, other functionalities found
    in conventional generating plants such as power system stabilization and participation in
    automatic generation control would require further studies and discussions between grid
    operators and manufacturers.

Workshop Summary - Wind Delivery and Reliability                                 March 30-31, 2005

      The goal of the workshop was to understand the issues associated with an increased
      penetration of wind energy in the North American market. It also presented a North
      American perspective and some European experience in wind integration. It brought
      together the major stakeholders, including utilities, manufacturers and regulators, in a
      forum dealing with the specific issues of wind energy delivery and reliability. The
      workshop was organized jointly by AWEA, CanWEA and UWIG, with participants from the
      United States, Canada and Europe.
      The specific objectives were to discuss the issues surrounding wind energy delivery and
      reliability, to promote solutions to wind integration, particular with hydro, and to create a
      forum for standardization efforts in the area of grid codes. Related issues included
      understanding wind energy variability and its relation to energy demand, the potential of
      hydro-electricity to manage the variability, meeting the transmission system operator
      requirements and understanding the wind energy conversion systems as they relate to
      grid integration.

2.1   Welcome & Opening Session
1.a   Opening Remarks
      Donna Cansfield, Parliamentary Assistant to the Ontario Minister of Energy – Wind
      The efforts of Ontario in the energy sector were summarized, including the establishment
      of conservation targets, the removal of barriers to the integration of renewables, including
      wind energy, and government commitments to renewables. It was indicated that Ontario
      consumes 50 % more energy than New York, and that wind could provide some of the
      new capacity required in the future. The significant government commitments to
      supporting renewable energy were summarized among which: making available crown
      lands for wind farm installations, supporting the proper mapping of wind resources, net
      metering, supporting educational programs (Waterloo, the Ontario University of
      Technology), and providing tax incentives for clean and renewable energy and capital tax
      exemptions. Green and renewable energy are also part of the answer to the growing
      problem of air pollution.

      Interconnection and Wind Integration - The State of the Debate
1.b   Randall Swisher, American Wind Energy Association
      The activities of AWEA in support of wind energy and the information sharing in this area
      were summarized. It was noted that there is a rising interest in wind energy, and that wind
      is becoming a significant market, with 20 GW expected to be installed by 2010 in the US
      (representing, a potential $100 billion market. However, there are barriers to
      implementation, including inconsistent policies (the on and off wind production tax credit),
      the obstacles and opposition to the siting of wind farms, transmission constraints and
      operational rules and practices. AWEA was holding discussions with FERC on grid codes
      and interconnection requirements, and with environmental activitists to gain a wider
      acceptance for wind energy. Transmission constraints included locations of wind farms far
      from loads and transmission network congestion. Solutions were proposed: upgrading
      transmission and adding capacity, curtailment of firm contracts to free up transmission

Workshop Summary - Wind Delivery and Reliability                               March 30-31, 2005

      capacity, identification of bottlenecks and required upgrades, and, on the longer term, the
      development of a large transmission infrastructure for serving wind-rich regions.
      Regarding operational rules, AWEA was developing a grid code proposal, which was
      consensus based and was proposed to entities such as NERC, which has established a
      wind task force. Work was also underway to define with the concerned entities, including
      RTOs and ISOs, appropriate market rules to facilitate the integration of wind energy.

1.c   Robert Hornung, Canadian Wind Energy Association
      It was pointed out that Canada was only beginning to realize the potential wind could
      have to make a significant contribution to the energy portfolio. There are 2 GW of wind
      energy planned or under construction and CANWEA is targeting 10 GW for 2010,
      covering 4.5 % of the total electricity demand. The main challenge is the transmission
      grid, which is mostly controlled at the regional level. Other issues related to wind that
      need to be mastered are its variability, and the associated storage issues, including hydro
      storage, and meeting grid operator concerns. The workshop addressed these and related

2.2   Wind Integration: Stability and Grid Operations
2.a   An Overview of Studies Undertaken in North America
      Robert Zavadil, EnerNex Corporation - Utility Wind Interest Group
      There are two issues that need addressing with wind energy: interconnection and
      integration. Interconnection deals with technical issues, including abnormal conditions
      such as short circuits and loss of network elements. Interconnection issues are more
      engineering-oriented than policy-oriented. Integration addresses the broader issues of
      economics, such as the cost of production, the interaction with demand and generation,
      transmission congestion and the operation of control areas. The relationship of both these
      issues to system reliability is discussed. Regarding interconnection, the main issue is the
      ride through capability and the ability of the wind turbine generator to remain connected
      following major disturbances, and grid disruptions. In the area of integration, the concern
      is impact on grid operation. There are a number of studies dealing with low voltage ride
      through, reactive power management and the impact of large-scale addition of wind
      energy on the grid. Studies show that integration costs are modest for penetration levels
      up to 15 %. Curtailment, if implemented would have a significant impact on the

2.b   Impacts of Large-Scale Wind Generation on the Ontario Electricity System
      Michael Brower, AWS Truewind, LLC
      The objectives of this study were to assess the potential impacts of wind energy
      production on the electricity system of Ontario and the coincidence of wind plant output
      and load consumption in different time intervals. The approach was to simulate the
      integration of 2000 MW of wind in 6 zones and at 17 sites. Samples at 10 min intervals
      were taken, concurrent with 10 min load data. Wind energy was converted into electrical
      energy using a 10-12 % loss factor and with a capacity factor varying from 25 to 40 %.
      Conclusions from the study were the following: (a) there was a reasonable match in
      seasonal patterns between wind and load, particularly in summer and winter patterns; (b)
      the match was not good in the daily patterns. Capacity factors were found to be higher in
      winter (typical maximum of 45 %) than in summer (typical maximum of 18 %). Geographic

Workshop Summary - Wind Delivery and Reliability                               March 30-31, 2005

      diversity may be important. In general, the addition of wind energy in the electric grid
      resulted in lower than expected net power variability. Although load following
      requirements were small, the use of other complementary forms of energy was required.

2.c   The Effects of Integrating Wind Power on Transmission System Planning,
      Reliability, and Operations of the New York State Power System
      Nicholas W. Miller, GE Energy
      The presentation summarized the results of a study investigating the impact of the
      addition of 3000 MW, or 10 % of existing generation, on the NYISO grid (NYSERDA
      Phase 2 Report: Operations and Stability Impacts). Sites were selected to minimize
      impact on the transmission system, and to ensure geographic diversity of the installations.
      Issues of concern included the impact on the ISO operations and on the reliability of the
      grid, and the economics of wind generation. The study used a quasi steady state analysis
      based on 3-hour study periods. Steady state results showed that, consistent with a
      statistical analysis: (a) regulation increased by only a small amount (163 MW maximum);
      (b) load following requirements increased slightly (24 MW); and (c) power ramp rates
      were not an issue of concern. The stability analysis showed that there was no negative
      impact on: (a) the second to second automatic generation control (AGC) response; (b)
      transient stability and damping characteristics of the system. The study concluded that
      with the addition of wind generation: (a) there were few instances where existing units
      could not properly perform load following; (b) no economic penalty associated with
      dispatching need be considered; (c) there was no economic penalty associated with AGC
      and frequency control. In addition, the study showed that wind energy based on the
      chosen technology (doubly-fed induction generators) resulted in better voltage recovery,
      better damping and less voltage swings after faults, and that wind generation had no
      adverse effects on system stability. It also concluded that the low voltage ride through
      characteristic needed only to handle voltage down to 40 %.

2.d   Control Room View of Wind
      Mark Ahlstrom, WindLogics
      The Xcel Wind Integration Study (2004) reported on the consequences of integrating
      1500 MW (50 wind farms, 15 % system capacity) wind energy into the Xcel North System
      from the viewpoint of impact on reliability and ancillary services. It was based on a 10-
      minute time frame, and used hourly data for wind provided by a weather bureau, with
      meteorological models filling the gaps, and the system data provided by the system
      operator. Results showed that (a) the capacity value of wind was 27 % on average; (b)
      the power regulation increased by a small amount (7.8 MW); (c) the need for additional
      load following was negligible; (d) the cost of additional scheduling and unit commitment
      was small (4.4 $/MW). Results were consistent with the NYISO study. The conclusions
      were that: (a) there was no cost based justification for imbalance penalties for wind
      generation; (b) there was no significant impact of hour ahead scheduling, this would be
      handled by normal regulation; (c) there was a large value added in scheduling wind into
      next day unit commitment. Other issues raised included: (a) the advantages and
      disadvantages of centralized versus individual or local wind forecasting; (b) the need for
      proper forecasting for optimal scheduling in unit commitment and cost allocation; (c) the
      good forecasting technology and the reliability of the data; (d) the aggregation of wind
      turbine generators and of wind farms to reduce variability; (e) the need to integrate wind
      energy system into control room operations; (f) the necessity to forecast extreme events.

Workshop Summary - Wind Delivery and Reliability                                 March 30-31, 2005

      Control room integration was viewed as essential for making proper decisions and
      identifying cost impacts. Proper tools need to be developed for operators.

2.3   Wind Integration: European Approaches & North American Perspectives
3.a   Experience with Integrating Wind Energy in Western Denmark
      Paul-Frederik Bach, Eltra, Denmark
      Wind development in Denmark is driven by a national package of measures and has led
      to an installed capacity providing 16 % of energy. The issues related to the impact of a
      large penetration and its risks include: (a) the constraints on installed capacity that must
      be accommodated, particularly heat generation in CHP units; (b) the match between
      production and load; (c) the creation of an export market for residual wind power, the
      large thermal units having to continue operation; (d) the impact of extreme weather
      conditions, namely storms, leading to the shutting down of most wind generation and
      requiring manual reset; (e) the necessity for proper wind forecasting and weather
      prediction. The consequences of high penetration on the Danish system are: (a) the need
      to purchase regulating power and improve load following; (b) the imbalance of reactive
      power; (c) the debatable use of wind power capacity credits; (d) the strong fluctuation of
      wind energy prices on the spot market due to variability. High penetration raises the
      following issues: (a) the power system was not designed to accommodate high wind
      penetration, particularly from the system security viewpoint; (b) operation of the system
      and integration of wind energy rely on strong interconnections with neighbouring grids,
      namely with Germany, which may be required to provide balancing power. New system
      architectures must be considered for the future. The purpose and objectives of the new
      system would be to: (a) balance domestic supply and demand; (b) provide efficient grid
      control and operator control; (c) provide a defence against blackouts, a feature that does
      not exist in the present configuration. The situation raises the following issues associated
      with an expansion of wind energy and related to the relationship between wind energy,
      markets and demand: (a) the role and control of CHP plants; (b) grid stabilization; (c) the
      market for residual power and the relationship between markets and wind generation; (d)
      the consequence of wind generation on the cost of electricity, due to the increasing
      complexity of the system.

3.b   Wind Integration – US Perspectives
      Hank Courtright, Electric Power Research Institute (EPRI)
      Maximizing the value of wind power must address the following issues: (a) the location of
      wind farms, usually installed far from loads and connected to weak grids, with the
      possible exception of offshore installations; (b) the need to provide ancillary services such
      as spinning and non-spinning reserve; (c) the combination of wind with hydroelectric
      generation and storage and the associated balancing and shaping services. Examples of
      effective solutions include: (a) the shaping service provided by the Bonneville Power
      Administration (BPA), which offers a one week hold and re-dispatch service; (b)
      compressed air storage plants (CAES), which in addition to capacity and shaping
      features, can provide renewable energy credits for the recovered energy; (c) the
      combination of wind, combined heat and power (CHP) and hydro on weak grids and

Workshop Summary - Wind Delivery and Reliability                                March 30-31, 2005

      requiring storage, such as the Hawaiian system. Research needs for more effective wind
      power integration are summarized, including wind turbine models, storage, market
      mechanisms, integration of wind and load forecasts with economic dispatch, and dynamic
      thermal rating of transmission.

3.c   Wind Integration: a Canadian Perspective
      Rick Patrick, SaskPower / Canadian Electricity Association (CEA)
      Elements of a Canadian perspective include: (a) the growth in energy demand (1.5 % per
      year); (b) the expected older plant retirements and the resulting energy shortfall; (c) a
      more sensitized public. The questions that need answering include: (a) the economics of
      total electric power generation and its relationship to the environment; (b) the regional
      context and resources; (c) the transmission grid and the absence of economic incentives
      and national drivers to expand it; (d) the economics of wind, particularly considering
      increases in the price of gas; (e) the role and need for wind incentives (such as the
      WPPI). CEA’s roles include: (a) carrying out system integration studies; (b) working in
      collaboration with CanWEA to develop a safe, reliable and low cost wind energy supply.

2.4   Wind / Hydro Integration
4.a   Weakly-Interconnected Systems: Britain and Ireland
      Nigel Scott, Garrard Hassan
      Great Britain is a self-contained system, with essentially a north-south energy flow and
      with limited interconnections with the continent. To make the system more integrated and
      functional, the existing three RTOs will be merged into one. Features of the electricity in
      British market include: (a) a bilateral trading scheme operating on an hourly base; (b)
      system charges based on location to reduce the North to South flow, generation being
      located mainly in the North and loads in the South; (c) green certificates, providing 3
      times the standard rate for green power. Significant expansion of wind energy is planned,
      from 0.9 GW to 26 GW by 2020. Issues under consideration include: (a) dealing with
      transmission constraints, which will no longer be allowed; (b) grid codes; (c) the impact of
      wind on security and services; (d) generation scheduling; (e) integration and integration
      costs. Studies show that: (a) there will be no significant integration costs (5 % added cost
      for a 20 % penetration); (b) some transmission reinforcement will be required; (c) the cost
      of electricity is expected to increase by 5 %; (d) proper wind forecasting is needed; (e)
      new grid codes need be developed to address security and grid control issues.

4.b   Hydro-Québec Generation’s Studies on Wind Generation Integration
      Roger Lambert, Hydro-Québec
      Hydro-Quebec plans to integrate at least 1500 MW in the next 10 years. Integration
      studies are being initiated that will take into account wind variability and forecasting
      uncertainties, daily and seasonal cycles, the electric system characteristics and
      constraints, including load and conventional generation. The required data include
      meteorological data, wind turbine characteristics and electrical system data. The technical
      studies will examine the available resources, the intermittency of the resources, and the
      capacity value of wind. Impacts will be considered in different time frames: frequency
      regulation (less than 10 minutes), load following and unit commitment (less than one
      week), unit commitment and wind-hydro integration (less than one year) and long term
      generation planning. The general approach and methodology are based on data

Workshop Summary - Wind Delivery and Reliability                                 March 30-31, 2005

      acquisition and analysis using statistical methods. Storage and shaping services will be

4.c   Manitoba Hydro Work on Wind
      Ed Wojczynski, Manitoba Hydro
      Manitoba need not rely on wind energy to meet its needs, as the province can count on
      firm surpluses for export up to 2020. However, it will consider wind as a commercial
      opportunity for long-term firm contracts. Variability is therefore an issue, as well as
      integration costs, including shaping services, regulation and load following. Studies are
      carried out to evaluate wind integration benefits, namely accredited capacity, and wind
      energy and integration costs, including shaping and firming, regulation and load following,
      and market and risk costs. The availability of hydroelectric resources for shaping services
      will dictate the amount of wind that can be integrated. The studies will also consider
      issues such as sites, transmission, forecasting, and interconnections.

4.d   Current Issues in Wind and Hydro Forecasting and Assessment
      Kenneth Westrick, 3TIER Group
      Extensive studies show that in forecasting wind and water, the following must be
      considered: (a) time frames and horizons; and (b) atmospheric weather pattern
      predictability, taking into account the fact that accuracy drops with time. To counter this
      present limitation there is a need for stochastic models and for sensitivity analysis. In the
      specific case of wind and hydro, the issue is correlation of effects in a given region:
      typically for hydro, it appears that the distance required for uncorrelated is 1500 km. In the
      case of wind, the studies have not yet been conclusive, but it appears less dispersion is
      required to obtain uncorrelated behaviour. Spectral analysis on data reveals the following
      cycles: yearly, 365 days, 911 days (ENSO cycle) and 3645 days (decadal). Wind and
      hydro both exhibit oscillatory behaviours, but with different time cycles. Studies reveal that
      wind has less long-term volatility, meaning it is more predictable and accountable on the
      long term, whereas hydro has less short-term volatility. Proper forecasting is therefore
      essential to evaluate the long-term value of a resources such as wind and hyfro. On the
      short term, the main issue is integration into the electric grid, an engineering issue. On the
      long term, it is essential that the risk assessment of the resource be properly done and
      this is largely site dependent. Proper placement of wind generation is therefore essential
      to maximize economic benefits. Wind and hydro integration must also consider the wind
      availability under low water conditions. Another issue is climate change that can have a
      significant impact on the long-term availability of a resource. In the case of hydro, this
      may lead to a need to change hydro reservoir dimensions. Significant changes in weather
      patterns are expected over the next 50 years. Extensive studies conclude that: (a) wind
      energy does provide additional energy capacity; (b) forecasting of both wind and hydro
      resources is important; (c) site selection must be considered.

2.5   Reliability Interconnections with Wind
5.a   Capacity Value and System Reliability Contribution from Wind Power
      Michael Milligan, National Renewable Energy Laboratory
      The contribution of wind to system reliability can be expressed in terms of capacity value.
      The capacity value is a way of measuring relative plant contributions to overall system
      reliability. It should be noted that total reliability may be expensive and a probabilistic,

Workshop Summary - Wind Delivery and Reliability                                 March 30-31, 2005

      rather than a deterministic approach, may be more appropriate. Capacity value or credit,
      also called effective load carrying capability (ELCC), is calculated using empirical data.
      The general method can be applied to wind generators. Simplified approaches can be
      used and were presented. However capacity values thus obtained may not match the
      ELCC. It is concluded that these methods should be used with caution, and although
      ELCC is data intensive, it is the more precise.

5.b   Issues and Results with Wind Integration
      Dennis Woodford, Electranix Corporation
      Wind and hydro integration should be considered a planning issue, and take into account
      the low voltage ride through behaviour (LVRT), the capacity of existing transmission lines,
      and the intermittency of wind. A good LVRT characteristic makes interconnection with the
      grid easier. In a deregulated system, transmission capacity is tied up in existing long term
      transmission contracts and available transmission capacity (ATC) may be non-existent.
      ATC may be freed for renewables by allowing limited curtailments. Intermittency can be
      handled by means of the storage and shaping service provided by hydro plants. The
      Bonneville Power Administration (BPA) offers such a service. This service is or could be
      made available in other jurisdictions such as Hydro-Quebec, BC Hydro, Manitoba Hydro.
      If power reversal is available in hydroelectric plants, these can be made more efficient by
      using pumped storage.

5.c   AESO Wind Interconnection Standard
      John Kehler, Alberta Electric System Operator
      Alberta plans to install a significant amount of wind, 760 MW (6 %) by 2006. It therefore
      had an interconnection standard developed specifically for wind. Extensive studies were
      carried out, including wind generator operation under network faults and the impact on
      voltage regulation. The proposed new specifications addressed the issue of voltage ride
      through and covered reactive power needs, and voltage and frequency operational limits.
      The intent of the new grid code is to consider the possibility of treating wind generation in
      a manner similar to other generation to the extent warranted. The power system stabilizer
      function for example was not deemed necessary. The grid code also covers monitoring,
      testing and validation. In addition, it requires that proper models be available to assess
      system impacts. Studies that need be completed include a wind variability study and an
      operations impact assessment study. Power management was considered optional at this
      time and its need will be reassessed after the variability studies are completed.

5.d   Summary of FERC, WECC, AWEA Proposals on Low Voltage Ride Through,
      Streamlining Interconnection Process
      Jim Caldwell, PPM Energy
      The proposed LVRT curve is based on the German E.On standard, as it was felt that this
      curve was suitable for the typical grids, and that manufacturers were already designing
      their wind turbines to that standard. The proposal has been submitted for approval. Other
      issues of interest included voltage support and control and reactive power requirements.
      The principle that wind generators are expected to contribute to the supply of reactive
      power was accepted. It was however also acknowledged that wind systems are a high
      cost producer of reactive power and that there may be a justification for compensating the
      production of reactive power. The proposed power factor range was set to be between

Workshop Summary - Wind Delivery and Reliability                                March 30-31, 2005

      0.95 leading and 0.95 lagging. Other issues included generation management
      (instantaneous output limits, ramp rate control, staggered high wind speed cut out). It was
      felt that there was a need to codify all requirements in one document.

2.6   Incentives
6.a   Update on the Production Tax Credit (PTC) Extension
      Jaime Steve, American Wind Energy Association
      The PTC is due to expire on December 31, 2005 and it is of utmost importance for the
      further success of wind energy implementation that it be extended (if possible long term, 3
      years) or renewed. The options available to Congress were explained. It was also pointed
      out that successful integration of wind required the availability of transmission capacity,
      the PTC and credits for renewables. Entities and individuals were encouraged to write to
      Congress in support of the PTC extension.

6.b   The Wind Power Production Incentive (WPPI) - Update
      Robert Hornung, Canadian Wind Energy Association
      The value of the WPPI is set at 1 ¢ / kWh for a period of 10 years. This ensures long-term
      support for wind, however its value is less than the PTC. The WPPI has been
      instrumental in significantly increasing the amount of wind energy installed or planned.
      However, the total amount of the subsidy is fixed and it is expected that it will be fully
      allocated by 2005-2006. CanWEA was seeking an allocation extension to allow the WPPI
      to cover projects to 2010. In addition, the WPPI should be combined with other incentives.
      However it is ensure if a project accepting the WPPI would be allowed to keep the
      environmental attributes associated with renewable energy.

2.7   General Session
7.a   Wind Integration Experience
      Garry Brown, VP, R & D, New York State Independent Operator
      Wind is a new technology that does not have the same attributes as other sources of
      energy. The issues include the amount of wind that can be accommodated (for example
      10 % or 3300 MW), the impact on the reliability of the electric supply, the impact on
      markets, and the non-discriminatory integration of fluctuating resources such as wind. It is
      accepted that there is a need and willingness to diversify energy sources. Options
      however are for all practical purposes limited to wind, part of the renewable portfolio
      standard. Issues and options of integrating wind include: (a) locational pricing; (b)
      imbalance penalties, considering present rules based on conventional power dispatch;
      and (c) wind capacity and its impact on markets. The GE study concluded that there were
      no technical or market difficulties in integrating 3300 MW of wind energy if consideration
      was given to specific wind issues and proper market rules were designed. Key findings of
      the study concluded that: (a) the addition of wind may increase congestion; (b)
      forecasting is important, with the possible need for centralized forecasting; (c) voltage
      requirements were met and better voltage regulation could be demonstrated in a number
      of cases; (d) single contingency reserve could be met; (e) market rules may have to be

Workshop Summary - Wind Delivery and Reliability                               March 30-31, 2005

      Wind Interconnections: Strategies and Issues
7.b   Strategies for Interconnection
      Géza Joós, McGill University
      Wind energy technologies and their impact on the grid were summarized and the
      operational considerations and limitations presented. Interconnection issues and network
      considerations were discussed. Existing standards and grid codes were presented
      including general requirements, and the specific requirements for distribution and
      transmission level interconnections. Present and future generic wind turbine generator
      characteristics were classified. Typical grid code requirements, including low voltage ride
      through and the AWEA proposal were presented, with reference to existing Canadian and
      foreign grid codes. Additional impacts of wind on the grid operation were mentioned. It
      was argued that standardization of grid codes, as undertaken by AWEA, could be
      beneficial to all parties involved in the interconnection of wind into the electric grid.

7.c   Managing Distribution-Level Wind Connections
      Jeff Smith, EPRI Solutions
      At the distribution level, the quality of the voltage was deemed to be the most important
      issue. This was quantified as short-term effects, namely flicker, and longer term effect,
      such as voltage regulation. Voltage regulation is usually reduced by means of switched
      capacitors (reactive power control) and transformer tap changers. Active voltage control
      yields a ± 2 % regulation, compared to a ± 5 % for conventional systems. Flicker,
      associated with wind variations, must be kept within the limits imposed by standards (IEC
      and others). Other issues discussed included protection coordination, anti-islanding and
      harmonic injection.

7.d   Wind Park-Level Approach to Interconnection
      Steve Saylors, Vestas Americas
      The issues associated with wind-park level compensation were presented, in particular as
      related to the differences between turbine based and external compensation. Although
      wind generation was previously exempt from having to provide voltage regulation, grid
      codes have evolved and there was a trend towards applying this requirement which
      applies to conventional generators. External compensation is based on the use of
      equipment such as the SVC and the STATCOM, which can be provided separately. This
      compensation, associated with SCADA equipment, can meet the interconnection
      requirements, including the low voltage ride through characteristic and grid voltage
      regulation. Other wind farm requirements may include power control and frequency
      regulation, controllable rate up and down power ramping, power curtailment, and spinning
      reserve provisions. SCADA systems improve grid interconnection operations.

2.8   Vendor Forum on Meeting Grid Voltage Standards - Technology Options
8.a   Meeting Grid Voltage Standards
      Steve Saylors, Vestas Americas
      Wind turbine generators offerings include grid connected induction generators and doubly
      fed induction generators. Grid integration requirements were discussed, including reactive
      power requirements and control, low voltage ride through (LVRT), and SCADA systems.
      LVRT requirements were compared for different types of generators and results obtained

Workshop Summary - Wind Delivery and Reliability                              March 30-31, 2005

      for specific installations were presented. The importance of improved SCADA systems for
      turbine and wind park control and data collection was stressed. It was indicated that
      models were provided for PSS/E, PSCAD/EMTDC and PSLF studies of transient and
      steady state operation.

8.b   Meeting LVRT, Voltage and Reactive Power Requirements
      Nick Miller, GE Energy
      It was indicated that so far the low voltage ride through (LVRT) requirements have been
      different from one installation to another. Available characteristics were discussed,
      including the proposed AWEA LVRT. Reactive power control and voltage regulation
      characteristics were presented. The implementation of voltage regulation at the individual
      wind turbine generator level and at the wind farm level was discussed. An example of a
      wind farm management system was given. It provided voltage regulation at the point of
      interconnection of the farm with the electric grid.

8.c   Under Voltage Ride Through Capability of ENERCON Wind Turbines
      Stephan Wachtel, ENERCON
      Offerings by the manufacturer are based on the same structure and components, scaled
      in power to meet customer demands. The technology is based on direct drive (gearless)
      variable speed synchronous generator with full power electronic converter control. Issues
      presented included: power quality characteristics (flicker and harmonics to IEC
      standards), reactive power control (unity power factor operation at the turbine level) and
      voltage regulation (at the wind farm level using a SCADA system), meeting the LVRT
      requirements, testing (for response to three phase faults), modeling (PSS/E, block
      diagram, Simulink models), and voltage regulation coordination (including control of on-
      line tap-changers). Extensive test and measurement results were presented.

8.d   Wind Grid Voltage Compensation Systems
      Gerald Keane, S&C Electric Company
      Substation-based compensation systems required for some types of turbine generator
      technologies connected directly to the grid to meet interconnection requirements were
      presented. Advanced systems included dynamic var compensation using mechanically
      switched capacitors, thyristor-based switched capacitors and IGBT inverter based
      systems (D-STATCOM). Operational features were described. A newer product
      specifically designed for wind farm, the Electronic Shock Absorber (ESA), provided both
      voltage and frequency regulation. Energy is stored in ultra-capacitors, which allow real
      power to be supplied or absorbed for short periods.

8.e   Meeting Grid Voltage Standards
      Bud Kehrli, American Superconductor
      A range of equipment designed to meet grid code requirements including power factor,
      dynamic voltage regulation and low voltage ride through (LVRT) were presented. These
      included dynamic var devices and switched shunt devices. It was shown, by means of
      practical examples, that meeting grid requirements may demand the installation of
      additional compensation equipment. The consequences of the voltage regulation
      requirements on mechanical switches were discussed. It was pointed out that the
      proposed LVRT standard (based on the German Eon Netz standard) did not provide

Workshop Summary - Wind Delivery and Reliability                              March 30-31, 2005

      comparable treatment for wind farms as for conventional generation, and should consider
      fault events with more realistic locations, depths and durations. This had a direct impact
      on the cost and complexity of the installations.

8.f   A Wind Generation Developer’s Perspective
      Mark Scher, MSE Power Systems
      The issues associated with interconnection to the electric grid were addressed, namely
      keeping the generation on line during faults, and supporting voltage in the case of remote
      transmission lines. Dynamic and non-dynamic reactive power compensation systems
      were described. General questions raised included the type of reactive power required,
      namely conventional capacitor or dynamic compensation, and the necessity to tailor the
      solution to the exact needs, in order to control costs.

Workshop Summary - Wind Delivery and Reliability                                 March 30-31, 2005

3.1   Comments - Summing up the issues – an opinion
      Size matters – In order for wind to be properly integrated and take advantage of diversity,
      control areas may have to be made as large as possible, and some consolidation may be
      required. This will make dynamic scheduling and generation control easier.
      Market structure matters more than the physics – Tariffs must be adjusted to take into
      account the nature of wind energy and the power imbalances it creates. Special rules are
      needed, particularly in high wind penetration regions. The system as a whole should be
      balanced and not individual systems or transactions. There is a belief that market forces
      will naturally result in an adequate system operation.
      System flexibility matters most – More flexibility should be introduced into the system
      to accommodate new technologies and reward new initiatives. It should be pointed out
      that the base load generation is subsidized and that expensive fuel may be burned during
      low demand periods. Removal of subsidies should be considered. Equipment design
      trends should be examined and more flexibility and larger design margins built into
      generating units if regulation and balancing functions are implemented.
      Non-traditional transmission planning – This approach should be considered for
      complete systems, not only those that integrate wind.
      Related issues – Important associated issues include information transmission and
      sharing, education and communication.

3.2   Question periods – issues raised - opinions
      Role of dc transmission – There are a number of dc transmission systems. For offshore
      wind, dc transmission can be considered for distances greater than 60 km. Ground return
      is allowed for short periods (30 min).
      Incorporating flexibility – A market structure rewarding initiatives is required.
      Capacity credits and impact of penetration – There is no standard answer to the
      relation between impact and levels of penetration. Calculations must be done for a
      specific system.
      Testing for compliance to LVRT specification – Validation based on specific system
      design should be carried out.
      Evolving from an accountant controlled to an engineer controlled industry – Keys to
      evolution of the electric grid include universities and the revival of power engineering
      education. It should also be easier to integrate new technologies, and there should be
      better incentives for wind and clearer market signals.
      Market for reserve and ancillary services – Establishing this market is difficult,
      particularly for reactive power. The real power and energy markets are already difficult
      enough to handle.
      Emission reduction credits – These should remain with the wind farm developer.

Workshop Summary - Wind Delivery and Reliability                               March 30-31, 2005

     Incentives for incorporating wind forecasting in a wind farm – These are the
     penalties to which the producer may be subjected. In addition, better forecasting may
     allow the wind farm to behave in a manner similar to other generators.
     Justification for building wind farms when imports are available – The tie line may
     have limited capacity and congestion may become an issue. In addition, it makes good
     economic sense to have generation close to the load, as this increases the reliability of
     the grid and reduces its vulnerability to transmission line loss.
     Protection – Mechanisms of implementation are discussed including signals for breaker
     Coordination of active voltage controllers – Adjustments and proper design of
     controllers should remove the possibility of voltage instability associated with controllers
     with conflicting settings and targets.
     Impact of full converter systems on the grid – There is no radical difference with the
     more conventional grid coupled induction or synchronous generators under most
     Reactive power requirements for underground cables – More capacitance and less
     reactance alter the reactive power requirements for voltage regulation and power factor
     Dynamic and non-dynamic reactive power equipment – Dynamic reactive power
     solutions, usually associated with the power electronics incorporated in the wind turbine,
     dominate the market.
     Impact on voltage of dynamic and non-dynamic reactive power equipment – The
     effect is more pronounced on weak systems and is minimal for strong power grids.
     Necessity for dynamic compensation – This should be used when needed. It should be
     noted that attempting to implement dynamic compensation using mechanical switching
     equipment might lead to premature equipment wear. However, studies of the operation of
     mechanical and electronic equipment reveal that both solutions give similar results. The
     number of mechanical switching cycles remains within equipment design specifications.

Workshop Summary - Wind Delivery and Reliability                                 March 30-31, 2005

4.1   Summary of Topics and Audience
      The workshop has addressed most of the important issues in wind energy delivery and
      reliability, and presented the point of views of a diverse cross section of the stakeholders,
      namely utilities, manufacturers, consultants, as well as government and interest groups.
      The approaches and positions taken by stakeholders from across North America were
      represented, as well as a sampling from Europe.
      The workshop has attracted a significant number of attendees from all sectors, including
      utility, regulators, industry, and academia. Attendance for all the sessions was very high.
      Most presentations stimulated questions from the audience.
      The scope of the workshop covered wind energy and reliability issues, and examined
      issues in terms of:
      -   geographic regions: North America and Europe;
      -   economic sectors: utilities, manufacturers, consultants, government;
      -   technical interests: electrical and mechanical engineering, weather forecasting;
      -   technical level and depth: generally understandable for an audience with technical

4.2   General Conclusions
      The main conclusions of the presentations and the ensuing discussions are the following:
-     Wind penetration levels of up to 20 % of the installed capacity will have a minimal impact
      on the more important aspects of the operation of the electric grid.
-     Storage, shaping and balancing are key issues for proper integration of wind energy and
      can be achieved using hydroelectric power.
-     Electric market structures and rules will have to evolve to allow a proper integration of
      wind energy and to give it a suitable place in the energy portfolio.
-     The grid integration issues at the wind turbine generator level are well understood,
      Manufacturers are willing to implement these requirements on their products, while
      insisting on the specific features of wind turbine technologies that make them different
      from conventional synchronous machine based power generation.
-     There are no objections in principle to harmonize the important aspects of grid codes.
      However, both utilities and manufacturers mention that uniform requirements for all
      aspects of interconnections and in all jurisdictions may not be desirable or realizable.
-     Wind farms could operate within limits in a manner similar to conventional generating
      plants. It is generally accepted that they need to and can participate in the voltage
      regulation of the electric grid at the point of connection. However, other functionalities
      such as power system stabilization and participation in automatic generation control
      would require further studies and discussions between grid operators and manufacturers.

Workshop Summary - Wind Delivery and Reliability                               March 30-31, 2005

4.3   Suggested Topics for Future Workshops
      Topics that could be considered in future workshops include:
      (a)   A more in-depth coverage of interconnection issues from the electrical engineering
            viewpoint, and the implications of grid codes on equipment design; this would be
            useful to engineers;
      (b)   A detailed presentation of the control system requirements to satisfy grid code
            requirements; this could be could be combined with (a);
      (c)   A discussion of the electrical and mechanical issues involved in the design of wind
            turbine generators and if possible a comparison between technologies offered; this
            could be generic and need not refer to a specific equipment manufacturer;
      (d)   A presentation of technical studies required for proper integration and the models
            required to carry out the studies;
      (e)   An explanation of capacity factors and the methods for calculating these factors;
      (f)   Detailed presentations on wind forecasting methods, and available tools;
      (g)   Market mechanisms, including the significance and use of environmental credits.


To top