Summary of Stability Base cases for TPL 001-4 Studies



Summary of Stability Base cases for TPL 001-4 StudiesThis document accompanies the base cases that have been posted to the ISO New England (ISO-NE) website under Planning Advisory Committee (PAC) materials on August 11, 2015 in order to coordinate with the PAC. The base cases will be utilized by ISO-NE to conduct studies needed to meet the NERC TPL-001-4 standard.Any text adjacent to the blue square brackets ([R1]) in the sections below refers to a specific requirement in the NERC TPL 001-4 standard. If a subsection has such a designation then the information within that subsection pertains to the stated requirement in the standard.The study base cases will cover the following:Five year out peak load case (2021) to assess the near-term stability impact to meet requirement [R2.4.1]Five year out off-peak load case (2021) to assess the near-term stability impact to meet requirement [R2.4.2]Study AssumptionsTo assess the stability performance of the New England system, two stability base cases have been developed. For the purposes of peak load stability analyses, the year five projection of the 90/10 summer peak demand level was used. For the purposes of off-peak load (light load) stability analyses, a 12,500 MW fixed light load was used. Source of Stability Models [R1]To represent the New England system topology, the Model on Demand (MOD) system was used. This included all ratings and topology in the MOD system as of May 22, 2015. For the external system outside New England, cases from the 2011 Multiregional Modeling Working Group (MMWG) case library were used. After creating 2021 peak and off-peak load cases from MOD, the cases were merged with an appropriate external case from ISO-NE’s 2014 August stability case library, to create final cases. Note that the New England system topology is identical for both steady state and stability cases. The stability model data from ISO-NE’s 2014 August stability library cases for 2018 Summer Peak and Light Load was used as the starting point for the stability case development. The stability model data for the new projects and changes to any existing facilities since August 2014 were incorporated.Existing Facilities [R1.1.1]Included in the base cases were all facilities that were in-service as of May 22, 2015 that were in the ISO-NE MOD system. Additionally, any updates to the model made between May 22, 2015 and June 30, 2015 were included. For existing facilities the ISO-NE MOD system relies on the NX-9 application to update the topology information. The NX-9 application includes the transmission data that is collected in accordance with ISO-NE Operating Procedure OP-16.The voltage and reactive schedules for transmission devices like autotransformers and dynamic reactive devices are based on voltage schedules that are published in the latest version (12/09/2014) of the ISO-NE Operating Procedure OP-12 Voltage and Reactive Control Appendix B.The following existing HVDC transmission facilities have models included for transient stability analysis:Phase IICross Sound CableHighgateAdditionally, the following existing dynamic reactive devices have models included for transient stability analysis:Chester SVCBarnstable SVCEssex STATCOMGlenbrook STATCOMSaco Valley Synchronous CondensersJay Synchronous CondenserGranite Synchronous CondensersKibby Dynamic VAR device (DVAR)Groton Dynamic VAR device (DVAR)Granite Dynamic VAR device (DVAR)As noted before, the stability model data from ISO-NE’s 2014 August stability library cases for 2018 Summer Peak and Light Load was used as the starting point for the stability case development. The model data for the existing generators was based on the 2014 August stability cases with the exception of units in Table 1 that had updates made to their models.Table 1 Stability Model Updates to the Existing GeneratorsGeneration FacilitiesDescription of Stability Model UpdateBridgeport Energy Increase in CT Updated model addedSkelton Station Modernization in MEUpdated model addedBethlehem unit in NHNew model added (previously no stability model was available) Cataract East unit in MENew model added (previously no stability model was available)Hemphill 1 unit in NHNew model added (previously no stability model was available)DG Whitefield unit in NHNew model added (previously no stability model was available)Northfield Mountain #2, #3, #4 Upgrades in MAUpdated model addedPt. Lepreau unit in NBUpdated model addedReEnergy Livermore Falls unit in MEUpdated model addedWestbrook Uprate in MEUpdated model addedBull Hill Wind in ME Updated model addedKibby Wind in ME Updated model addedRollins Wind in ME Updated model addedStetson Wind in ME Updated model addedSpruce Mountain Wind in ME Updated model addedSouth Meadow Unit #5 Automatic Voltage Regulator (AVR) replacement in CTUpdated model addedLong-Term Outages [R1.1.2]To determine long-term outages with duration greater than six months in the system conditions being tested in 2021, the June 2, 2015 Long-Term Outage Report was reviewed. There are no known long-term outages of transmission or generation facilities with duration over six months for the particular study periods, and hence no long-term outages were modeled in the base cases. This report is located in the Operations Reports section of the ISO-NE website: Topology Changes [R1.1.3]The May 18, 2015 Regional System Plan (RSP) Project List was used to select transmission projects included in the study base cases. All reliability upgrades (Table 1a and 1b) that were Proposed, Planned and Under Construction were included in the study base cases.Two future reliability upgrades that were on the May 18, 2015 RSP Project List were excluded – RSP 1269 and RSP 1280. These projects are discussed in section 1.1.12.In addition to the transmission topology changes, models for transient stability analysis were added for the following dynamic reactive devices:Ascutney Dynamic Voltage Support Device in VT (RSP 1615)Coopers Mills STATCOM in ME (RSP 1643) Stony Hill Synchronous Condenser (RSP 1578)The following dynamic reactive devices that will be retired as a part of Southwest Connecticut (SWCT) Transmission Upgrades were removed from the stability cases:Stony Hill dynamic devices in CTBates Rock dynamic device in CTA few other future dynamic reactive devices were added with the addition of certain approved queue generators. These are included in the section on generation assumptions.A data request for Local System Plan (LSP) projects that affect the BES was sent by the ISO on April 7, 2015 to Transmission Planners. This request was regarding projects that are not included in the RSP Project List. Each Transmission Planner used their discretion for which LSP projects should be included. A list of LSP projects that were provided by the Transmission Planners is included in Table 2. The list does not include LSP projects that have an associated project on the May 18, 2015 RSP Project List:Table 2 LSP Projects Included in StudyTransmission PlannerStateIn-service DateProjectSolutionsVELCOVT5/2016Rutland-Florence 46 kV InterconnectionClose normally open switches Marble St 237 and West Rutland 446. Reconductor 46 kV from West Rutland Tap - Marble St - Proctor Tap to Florence.VELCOVT5/2015White River SubstationAdd new 46/12.47 kV substation called White River Jct, transfer load from two existing substations (Norwich and White River), add new 46 kV capacitor bank (White River Jct), and reconductor 46 kV circuits (Taftsville - Quechee - White River Jct - Norwich Tap)WMECOMA9/2015UMASS-Amherst new substation (Tillson Farm)UMASS-Amherst will be constructing/owning/operating a new 115/13.8 kV substation with two 55 MVA transformers and a circuit switcher at AmherstThe transmission upgrades associated with the ISO-NE Interconnection Queue projects that had a Proposed Plan Application (PPA) approval as of April 1, 2015 were included in the base cases. These include dynamic devices associated with the following projects:QP 273 – Addition of a dynamic device (Synchronous condenser) interconnecting to the Keene Road 115 kV substation in ME QP 333 – Addition of a dynamic device (SVC) at the Detroit 115 kV substation in METhe basecases also include a detailed model for the 34.5 kV system in the Central Maine Power(CMP) service area, based on data submitted by CMP.Forecasted Load [R1.1.4]The 2015 CELT report was used to obtain the forecast loads for the peak load study case. A 90/10 summer peak load was used for the peak case and a fixed load level of 12,500 MW was used for the off-peak case. Table 3 summarizes the load levels used in each case.Table 3 Forecasted Load LevelsYearLoad Level (MW)2021 Peak32,6762021 Off-Peak12,500Load Power Factor Assumptions [R1.1.4]Load power factors consistent with the local Transmission Owner’s planning practices were applied uniformly at each substation. Demand resource power factors were set to match the power factor of the load at that bus in the model. A list of overall power factors by company territory can be found in the detailed load report in Appendix A: Load Forecast.Firm Transmission Service and Interchange [R1.1.5]No firm transmission service or interchange was modeled in the study cases because ISO-NE does not have advance firm transmission service arrangements. Please see “ISO-NE OATT Business Practices Section 1—OATT Transmission Services Summary” for details on the transmission services provided by ISO-NE.Generation Assumptions [R1.1.6]All existing generators that are included in the ISO-NE MOD system were included in the base cases. All future generators in the ISO Interconnection Queue that had a PPA approval as of April 1, 2015 were included in the base cases. One generator, QP 330, that had a PPA approval as of April 1, 2015 has since withdrawn, and this unit has been placed out of service in the base cases and will not be assumed online for any of the dispatches.In addition, generators that do not have a PPA approval but have an obligation through the Forward Capacity Market (FCM) were also included in the case. The following generators do not have PPA approval but are included in the base cases since they cleared in the FCM:QP 440 – Gas Turbine Expansion in New Haven county in CTQP 444 – Gas Turbines in Norfolk county in MAFor the off-peak stability basecases, the maximum capability of all the units was modeled at its Network Resource Capability (“NRC”) – winter rating. NRC Winter is the maximum amount of energy that a generator has interconnection rights to provide in winter. It is measured as the net output at the Point of Interconnection and represents the generator’s maximum output at or above 0 degrees Fahrenheit. For the peak load stability basecases, the maximum capability of the units was determined by the status of the unit in the FCM. Any unit that cleared in the FCM was modeled at its Summer Qualified Capacity rating and a unit not in the FCM was modeled at its Capacity Network Resource Capability (CNRC) rating (Maximum rating at 90 degrees or greater).The reactive dispatch for the peak load case of the existing generators is based on the peak load voltage schedules that are published in the latest version (12/09/2014) of the ISO-NE Operating Procedure OP-12 Voltage and Reactive Control Appendix B. The System Impact Study (SIS) data was used for any future generator.The reactive dispatch for the off-peak load case of the existing generators is based on the light load voltage schedules that are published in the latest version (12/09/2014) of the ISO-NE Operating Procedure OP-12 Voltage and Reactive Control Appendix B. The SIS data was used for any future generator.Photovoltaic (PV) generation is modeled in the peak load base case. A description of the PV generation modeling is included in section 1.1.11.No PV generation is modeled in the off-peak load case since the load is assumed to be fixed at 12,500 MW of New England load.The stability model data from ISO-NE’s 2014 August stability library cases for 2018 Summer Peak and Light Load was used as the starting point for the stability case development. The stability model data for the new projects and changes to the existing facilities since August 2014 were incorporated into the original stability data. Table 4 shows some of these major updates to future generation projects or future updates to existing generators:Table 4 Stability Model Updates to Future Generation Projects or Future Stability Model Updates to Existing Generators Generation/Transmission FacilitiesDescription of Stability Model UpdateGas Turbine Expansion in New Haven county in CT (QP 440)New model addedGas Turbines in Norfolk county in MA (QP 444)New model addedCombined Cycle Expansion in Fairfield county in CT (QP 412)New model addedBlue Sky West Wind in ME (QP 333)New model addedPisgah Mountain Wind in ME (QP 349)New model addedPassadumkeag Wind in ME (QP 357)New model addedPotter Repowering in MA (QP 438)New model addedWyman Hydro Uprate in ME (QP 366)Updated model addedNorthfield Mountain #1Upgrade in MA (QP 196) Updated model addedJericho Wind expansion in NH (QP 415-437)Updated model addedCanton Wind Increase in ME (QP300-406) Updated model addedSaddleback Wind Increase in ME (QP 287-407)Updated model addedThe below section shows some of the major stability models that were removed from ISO-NE’s 2014 August stability library cases for 2018 Summer Peak and Light Load. Table 5 shows the models that were removed due to the fact that some of the future projects were cancelled from the RSP Project List or withdrawn from the ISO-NE Interconnection Queue List. Table 6 shows the models that were removed due to the upcoming retirements before the study year of 2021.Table 5 Stability Model Updates for Withdrawn/Cancelled ProjectsGeneration/Transmission FacilitiesDescription of Stability Model UpdateDeerfield Wind in VT (QP 276)Model removed (Withdrawn Project)Bowers Wind in ME (QP 327)Model removed (Withdrawn Project)Fair Haven Biomass in VT (QP 330)Model removed (Withdrawn Project)Northern Pass HVDC in NH (QP 348)Model removed (Cancelled Project per March 2015 RSP Project List) Table 6 Stability Model Updates for Upcoming RetirementsGeneration/Transmission FacilitiesDescription of Stability Model UpdatePPL Great Works - Red Shield in MEModel removed (Effective Retirement Date of 6/1/2018)Brayton Point Units (#1, #2, #3, #4) in MAModels removed (Effective Retirement Date of 6/1/2017)Brayton Point Exciter Model Replacement (QP 347) in MAModels removed (Effective Retirement Date of 6/1/2017)Mount Tom Unit in MAModel removed (Effective Retirement Date of 6/1/2018)Norwalk Harbor Units #1, #2, #10 in CTModel removed (Effective Retirement Date of 6/1/2017)Medway Diesel unit in MEModel removed (Effective retirement date of 6/1/2017)Wallingford Refuse unit in CTModel removed (Effective retirement date of 6/1/2018)PPL Veazie unit in MEModel removed (Effective retirement date of 6/1/2017)Bar Harbor Diesel unit in MEModel removed (Effective retirement date of 6/1/2017)Brayton Point Diesel units in MAModel removed (Effective retirement date of 6/1/2017)Claremont unit in VTModel removed (Effective retirement date of 6/1/2018)Power system software tool PSS/E’s “GNET” function is typically used during stability simulations to convert the generators to negative loads if they do not have a dynamic model available or have insufficient dynamic model data. There were certain generators GNETed (i.e. converted to negative loads during simulations) in ISO-NE’s 2014 August stability library cases for 2018 Summer Peak and Light Load. Note that the units that were GNETed are smaller than 20 MVA for the unit and 75 MVA for the plant satisfying the NERC requirement. As a part of this NERC TPL stability case building, the following two future generators were added to the list of GNETed generators:Block Island Wind Farm in RI (QP 405): This is a 30 MW future offshore wind farm that will be interconnected in the town of New Shoreham in Block Island, RI. Due to insufficient stability data, this unit was GNETed in the stability cases. No significant impact is expected to the stability analyses for the transmission system as this unit will be interconnected to the sub-transmission network.Harrington Street PV Project in MA (QP 377): This is a 6 MW Photovoltaic Interconnection project in East Brookfield, MA. Due to insufficient stability data, this unit was GNETed in the stability cases. No significant impact is expected to the stability analyses for the transmission system due to its smaller size.Some important comments while using the stability cases are provided in Appendix D: General Comments on Stability Cases.Demand Resource Assumptions [R1.1.6]Demand resources (DR) were only modeled for the peak load case since the off-peak load case at 12,500 MW includes the impact of all DR. Demand resources that are modeled in the study can be divided into three categories:Passive DR that cleared in the Forward Capacity Market through FCA 9 (June 1, 2018 – May 31, 2019)Active DR that cleared in the Forward Capacity Market through FCA 9 (June 1, 2018 – May 31, 2019)2015 CELT Energy Efficiency (EE) forecast for the period beyond May 31, 2019.The following table shows the performance assumption for each category of DR that is included in the peak load base case. Real-Time Emergency Generation (RTEG) was not modeled in these analyses. RTEG that cleared in the FCM was not included in the reliability analyses because long term planning analyses should not be performed such that the system must be in an emergency state as required for the implementation of OP-4, Action 6. Performance Assumption (PA) factors are based on the expected performance of DR after receipt of instructions to dispatch. For Passive DR and forecasted EE the PA factor is assumed to be 100% since the DR programs are already installed and automatically reduce load. For Active DR, the PA factor is assumed to be 75%. This is based on historical response rates. Table 7New England Demand Resource Performance AssumptionsLoad LevelPassive DRActive DRForecasted EERTEGsSummer Peak (90/10)100%75%100%0%The details of the demand resources modeled is included in Appendix B: Demand Resources. A summary of the DR modeled across New England is provided in the tables below. Note that FCA9 DR refers to the sum of the qualified capacities of existing DR in FCA9 and the newly cleared DR in FCA9.Table 8 Passive DR Modeled in the Peak Load CasePassive DR?Study Year2021 PeakNew EnglandFCA 9 Passive DR (DRV)21355.5% Distribution Losses116Passive DR Modeled in Case2252Table 9 EE Forecast Modeled in the Peak Load CaseEE Forecast?Study Year2021 PeakNew EnglandEE Forecast for years 2019 - Study Year (DRV)6445.5% Distribution Losses35EE Forecast Modeled in Case680Table 10 Active DR Modeled in the Peak Load CaseActive DR?Study Year2021 PeakNew EnglandFCA9 Active DR (DRV) in New England4845.5% Distribution Losses+ 27Unavailable Active DR (25%)-128Active DR Modeled in Case383Market Solution Consideration [R1.1.6]In addition to the resources that cleared the FCM, the PV generation forecast was used to model PV generation in the summer peak load case. Similar to DR, the PV generation forecast was only modeled for the summer peak load case. The 2015 CELT PV generation forecast includes the PV generation that has been installed as of the end of 2014 and provides a forecast by state of the total PV (by AC Nameplate) that is expected to be in-service by the end of each forecast year for the next 10 years. As an example the 2015 PV forecast provides the PV that is in-service as of the end of 2014 as well as provides an annual forecast for the PV that will be in-service for end of 2015, end of 2016 and so on until the end of 2024. An availability factor of 26% was applied to the values from the PV generation forecast. The details of the PV generation modeling are included in Appendix C: PV generation modeling. Table 11 summarizes the PV generation modeled for the peak load case.Table 11 PV generation Modeled in the Peak Load CasePV generation Modeled as Negative LoadsStudy Year2021 PeakNew EnglandA - PV generation (nameplate) in New England1694B - 5.5% Reduction in Distribution Losses93C - Unavailable PV generation (A+B)*(1-26%)1322PV generation Modeled in Case as Negative Loads (A+B)-C465Explanation of Future Changes Not IncludedTwo transmission projects listed on the May 18, 2015 RSP Project List were not included in the study cases:RSP 1269: The project to add a new 345 kV breaker and two 25 MVAR capacitor banks at the Amherst substation in New Hampshire was not included because the Transmission Owner has raised concerns regarding the feasibility of implementing the Amherst 345 kV capacitor banks proposed in the previous New Hampshire/Vermont Solutions Study.RSP 1280: The project to add a new 115 kV line section 244 and upgrading line section 80 between Coopers Mill and Highland Substations was not included because the Transmission Owner indicated that the project may not move forward.Load Levels Studied [R2.1.1, R2.1.2 and R2.2]This section summarizes the net load level studied for each base case. The tables below exclude the Station Service loads which vary based on dispatch assumptions. Table 12: Net Load Modeled for 2021 peak load (Excludes Transmission Losses)CategoryLoad (MW)90/10 CELT Forecast32,676Non-CELT Manufacturing load in New England364Available FCA-9 Passive DR (modeled as negative load)-2,252Available FCA-9 Active DR (modeled as negative load)-383Available 2015 CELT EE Forecast for year 2021 (modeled as negative load)-680Available 2015 CELT PV Forecast for year 2021 (modeled as negative load)-465Net load modeled in New England (Excludes Station Service Load)29,260Table 13: Net Load Modeled for 2021 off-peak load (Excludes Transmission Losses)CategoryLoad (MW)90/10 CELT Forecast11,857Non-CELT Manufacturing load in New England364Available FCA-9 Passive DR (modeled as negative load)0Available FCA-9 Active DR (modeled as negative load)0Available 2015 CELT EE Forecast for year 2021 (modeled as negative load)0Available 2015 CELT PV Forecast for year 2021 (modeled as negative load)0Net load modeled in New England (Excludes Station Service Load)12,219Dynamic Load Model[R2.4.1]A dynamic load model was added to the peak load stability case to model the dynamic behavior of Loads that could impact the study area, including the consideration of the behavior of induction motor Loads. The breakdown of the load components for use with the WECC dynamic load model for New England is based on a report that was prepared by DNV-GL for Lawrence Berkeley National Laboratory. The report was titled “End Use Data Development for Power System Load Model in New England – Methodology and Results”. As a part of this report the New England load was characterized by:Summer peak and Spring Light Load hourSeven geographic areas - Maine, New Hampshire, Vermont, Connecticut, Rhode Island, Eastern Massachusetts and Western MassachusettsCustomer sector – Residential, Commercial and IndustrialElectrical end use categories specific to each customer sectorLoad components of the Western Electric Coordinating Council (WECC) The following bar chart summarizes the distribution of New England summer peak load in the seven regions into the 7 components of the WECC load model.At the time of commencing this analysis the ISO has not completed the benchmarking of the WECC dynamic load model for stability studies. Since the requirement R2.4.1 requires a dynamic load model that includes the behavior of induction motors, an equivalent dynamic model that would represent induction motors is being used for this analysis for the interim period before the WECC load model has been benchmarked and verified. The PSS/E standard complex load model (CLOD model) was used to perform this interim representation. The CLOD model has the following components: Large motor, small motor, discharge lighting, transformer saturation, constant MVA loads and remaining loads.Figure SEQ Figure \* ARABIC 1: CLOD Model ParametersThe WECC load model has the following parametersFigure SEQ Figure \* ARABIC 2: WECC Load Model parametersSince a direct translation of model parameters from the WECC load model to the CLOD model was not available, the ISO has made some assumption on how the parameters in the WECC load model are converted to the CLOD model. The following section describes the assumption made and the basis for each assumption.The following sections describes the different motor models in the WECC composite load model and the source is the WECC MVWG Load Model Report ver.1.0Motor A (MA) is representative of the across-the-line connected three-phase compressor motors used in air-conditioners and refrigerators. The driven load is constant torque. Motor B (MB) is representative of the across-the-line connected fan motors. The driven load is speed-dependent, proportional to speed squared. Motor C (MC) is representative of pump motors. The model data is similar to Motor B, except for lower inertia. Motor D (1-phase) is a representation of single-phase residential air-conditionersThe assumption is made that Motor A and Motor C represent small three-phase motors and Motor B represents large three-phase motors in the CLOD model. The following table shows a comparison of the parameters of Motor A, Motor B, and Motor C based on the WECC MVWG Load Model Report. The details can be found in the report, but in summary:MA and IA correspond to Motor A, MB and IB correspond to Motor B, and MC and IC correspond to Motor C. Based on the close inertia constants of Motor A and Motor C which is significantly lower than the inertia constant for Motor B, Motor A and C are categorized as small motors and Motor B is classified as a large motor. Note that since single phase motors are not explicitly modeled in the CLOD model, these single phase units are also modeled as small three-phase motors. In summary, the following conversion is used for the motor loads:Percentage of Large Motors = Percentage of Motor BPercentage of Small Motors = Percentage of Motor A + Percentage of Motor C + Percentage of Motor DThe next component in the WECC load model is the “Electronics load” component. The following is an excerpt from the WECC MVWG load model report for the behavior of electronic loads:Power Electronic loads - BPA tested Variable Frequency Drives, PCs and TVs in its laboratory. Power electronic loads behave as a constant power with respect to voltage and frequency. Power electronic loads operate close to unity power factor at the fundamental frequency. Electronic loads trip for low voltage condition at higher levels, usually 55% to 70%. The closest equivalent component to this performance in the CLOD model is that of the constant MVA load. It is a more conservative model than the behavior described above.The static component in the WECC load model consists of two different types of static loads – Constant current loads and constant impedance loads. The following excerpt from the WECC MVWG load model report describes what these loads consist of:A constant current component is representative of fluorescent lighting with electronic ballast, which is the majority of commercial lighting today. Ballast lights stay on until very low voltages, 10% to 15%. A constant impedance component is representative of resistive loads such as incandescent lighting, resistive water heating, resistive space heating, and resistive cooking. The discharge lighting model within the CLOD model has the following parameters:1. Real part of load as constant current.2. Imaginary part of load as the voltage raised to the power of 4.5.3. For voltage between 0.65 and 0.75 per unit, the real and imaginary loads are modeled as explained in the previous items (1. and 2.) with a linear reduction applied corresponding to the reduced voltage4. For voltage below 0.65 per unit, the lighting is assumed to be extinguished (i.e., no load).If we assume the constant current load under the WECC composite model is modeled as a discharge lighting component in the CLOD model we would see a difference in the performance between the two models depending on the system voltages. Since the CLOD discharge lighting model would disconnect at voltage below 0.65 p.u., and the WECC load would stay connected, for voltages below 0.65 p.u., the WECC load model would give a worse response than the CLOD model. However, at voltages between 0.65 p.u. and 0.75 p.u. the real power loads in the CLOD model will be lower than the WECC model and at voltages above 0.75 p.u. the real power loads would be the same for both models. The reactive power would be lower for the discharge lighting model in the CLOD model than the WECC model at all voltages below 1 p.u. However, given that a large part of New England has good load power factors (>0.99 p.u.) the impact of this reduced reactive and real power at lower loads is not expected to be significant. Further, the reduction of electronic loads at low voltages and the potential tripping of motor loads with low voltages is not modeled with the CLOD model and the additional conservativism in modeling motor loads and electronic loads is expected to offset the reduced constant current loads modeled in the CLOD model. In summary, the constant current load component in the WECC composite load model will be modeled as a discharge lighting load in the CLOD equivalent model.The only remaining component is the constant impedance component in the WECC composite load model. This component is modeled as a voltage sensitive load in the CLOD model where the load is set to vary by square of the voltage. There is no component in the WECC composite load model to represent transformer saturation. Hence the transformer saturation component of the CLOD dynamic load model is set to 0.In summary, the following methodology was used to convert the WECC composite load model parameters to the CLOD parameters:Percentage of Large Motors = Percentage of Motor BPercentage of Small Motors = Percentage of Motor A + Percentage of Motor C + Percentage of Motor DConstant MVA = Electronics Discharge Lighting = Constant Current LoadTransformer Saturation = 0Remaining Loads = Constant Impedance Loads with a Kp of remaining loads = 2 (for constant impedance loads)Two other parameters that are a part of the CLOD model but do not deal with the load components is the R+jX parameter that could be used to account for feeder impedances. The loads included in the basecases inherently account for distribution losses and hence no additional modeling of feeder impedances will be included in these basecases. It should be noted that for a majority of the New England loads the step-down transformers from transmission to sub-transmission levels (46 kV and below) is included in the basecases.Once we have established the methodology to convert the WECC composite load model parameters to the CLOD complex load model parameters the next step is to use the WECC composite load components from the New England end use load model to determine the actual components.New England loads consist of the following components:CELT load for the six New England statesPassive DR, Active DR and EE forecast for the six New England states modeled as negative loadsPV forecast modeled as negative loadsManufacturing loads in ME (Non-CELT)Station Service loadsThe first step in the load modeling was to eliminate all the negative loads associated with demand resources by reducing the load at the different buses by an equivalent amount. The next step was to reclassify all the New England CELT loads into the 7 regions defined in the New England end-use load model report. The CELT loads were reclassified into the following owners:Owner 60 – ME CELT loads net of DROwner 61 – NH CELT loads net of DROwner 62 – VT CELT loads net of DROwner 63 – Eastern MA CELT loads net of DROwner 64 – Western MA CELT loads net of DROwner 65 – Rhode Island CELT loads net of DROwner 66 – Connecticut CELT loads net of DRThe remaining non-negative loads are either Manufacturing loads in ME or station service loads across New England. To determine the components of the WECC load model that correspond to these loads the equivalent parameters of all the industrial load in New England was used. The following table summarizes the estimated Industrial Peak hour load component distribution. The total New England component distribution was used to assign the remaining manufacturing and station service loads. These loads were modeled under owner 67. The only remaining loads in the case are negative loads corresponding to PV loads and one negative load associated with equivalent generation in ME. Some limited number of loads in the case with zero real power output was given owner 69 and no dynamic loads were applied to these loads Since a large majority of the residual load is inverter based PV, all these loads are modeled using a ZIP load model with constant current characteristics for real and reactive power. . These reactive loads were modeled as constant impedance loads. The final table below shows the different components in the CLOD model for the different areas and loads in New England:RegionOwner IDLarge Motor (Motor A + Motor B)Small Motor (Motor C + Motor D)Discharge Lighting (Constant I)Transf. SaturationConstant MVA (Electronics)Other (Constant Z)KpMaine6012.09%42.62%11.86%0%15.95%17.48%2New Hampshire6111.97%42.19%12.78%0%15.74%17.32%2Vermont6211.22%45.61%12.15%0%15.13%15.90%2Massachusetts - East6312.27%47.58%11.59%0%15.82%12.75%2Massachusetts - West/Central6413.19%48.27%10.20%0%14.43%13.91%2Rhode Island6513.07%47.58%10.89%0%14.12%14.34%2Connecticut6611.77%45.43%12.11%0%17.60%13.08%2New England Manufacturing + SS loads (Non-CELT)678.93%54.06%1.99%0%20.19%14.84%2Negative load for generation in ME68Modeled as ZIP loads with constant current characteristics for P and QLimited number of loads with no real power component69Modeled as ZIP loads with constant current characteristics for P and QPV Gens as Negative loads94Modeled as ZIP loads with constant current characteristics for P and QLow Voltage Ride Through Capability [R4.3.1.2]Some of the units modeled in the stability basecases have relays that model generator tripping based on actual low voltage ride through capability. However, the majority of the units do not have protection systems modeled for this purpose. Figure 3-1 below shows the no-trip zone based on PRC-024-1. Since PRC-024-1 has an enforcement date of July 1, 2016 the ISO does not have specific low voltage trip set points for generators. A conservative assumption is made that all generators that do not have low voltage ride through based tripping modeled, the generator would trip for low voltages below the curve shown in Figure 3 below. An assumption was made that the protection relays would monitor the voltage at the high side of the GSU for each unit and trip if the high side of the GSU voltage is below the curve. The PSS/E generic model for Under/Over voltage generator trip relay (VTGTPA) was used. Four relays were used for each generator to simulate the four sections of the curve in Figure 3.Figure SEQ Figure \* ARABIC 3: PRC-024 Voltage No-Trip ZoneLine Relay Models [R4.3.1.3].The out-of-step scanning model (OSSCAN) will be used to scan the entire network for out-of-step conditions. With this model, whenever the apparent impedance as seen from either end of a line is less than the line impedance, the line is flagged as out of step. The model will only be activated after the last element cleared by protection is tripped. Additionally, the following specific relay models were included based on data provided by the transmission owners:PV20 Plattsburg - Sandbar Overload Mitigation System (OMS) relay model345 kV line 326 Special Protection System (SPS) thermal mode relay modelAppendix A: Load Forecast2021 Summer Peak Load2021 Off-Peak LoadAppendix B: Demand ResourcesThe following three tables indicate how each DR type is modeled in the peak load base case. The DR modeled in the peak load case by load zone is provided. The DR values that are released as a part of the FCA results and the 2015 CELT EE forecast have transmission and distribution losses (8%) included. For modeling in the study base case, only the distribution losses (assumed to be 5.5%) will be added to the DR since the transmission losses (2.5%) will be computed by the simulation software. Table 14: Passive DR Modeled in the Peak Load CasePassive DRStudy YearLoad Zone?2021 PeakConnecticutFCA 9 Passive DR (DRV)4255.5% Distribution Losses+ 23Passive DR Modeled in Case449MaineFCA 9 Passive DR (DRV)1685.5% Distribution Losses+ 9Passive DR Modeled in Case177New HampshireFCA 9 Passive DR (DRV)955.5% Distribution Losses+ 5Passive DR Modeled in Case100VermontFCA 9 Passive DR (DRV)1175.5% Distribution Losses+ 6Passive DR Modeled in Case123Rhode IslandFCA 9 Passive DR (DRV)1895.5% Distribution Losses+ 10Passive DR Modeled in Case199NEMA/BOSTONFCA 9 Passive DR (DRV)5275.5% Distribution Losses+ 29Passive DR Modeled in Case556SEMAFCA 9 Passive DR (DRV)2845.5% Distribution Losses+ 16Passive DR Modeled in Case300WCMAFCA 9 Passive DR (DRV)3315.5% Distribution Losses+ 18Passive DR Modeled in Case349New EnglandFCA 9 Passive DR (DRV)21355.5% Distribution Losses116Passive DR Modeled in Case2252Table 15: EE Forecast Modeled in the Peak Load CaseEE ForecastStudy YearLoad Zone?2021 PeakConnecticutEE Forecast for years 2019 - Study Year (DRV)1375.5% Distribution Losses+ 8EE Forecast Modeled in Case145MaineEE Forecast for years 2019 - Study Year (DRV)415.5% Distribution Losses+ 2EE Forecast Modeled in Case43New HampshireEE Forecast for years 2019 - Study Year (DRV)275.5% Distribution Losses+ 1EE Forecast Modeled in Case28VermontEE Forecast for years 2019 - Study Year (DRV)425.5% Distribution Losses+ 2EE Forecast Modeled in Case44Rhode IslandEE Forecast for years 2019 - Study Year (DRV)565.5% Distribution Losses+ 3EE Forecast Modeled in Case60NEMA/BOSTONEE Forecast for years 2019 - Study Year (DRV)1575.5% Distribution Losses+ 9EE Forecast Modeled in Case166SEMAEE Forecast for years 2019 - Study Year (DRV)855.5% Distribution Losses+ 5EE Forecast Modeled in Case90WCMAEE Forecast for years 2019 - Study Year (DRV)995.5% Distribution Losses+ 5EE Forecast Modeled in Case105New EnglandEE Forecast for years 2019 - Study Year (DRV)6445.5% Distribution Losses35EE Forecast Modeled in Case680Table 16: Active DR Modeled in the Peak Load CaseActive DRStudy YearLoad Zone?2021 PeakConnecticutFCA 9 Active DR (DRV) in dispatch zone Eastern Connecticut8FCA 9 Active DR (DRV) in dispatch zone Northern Connecticut28FCA 9 Active DR (DRV) in dispatch zone Norwalk-Stamford3FCA 9 Active DR (DRV) in dispatch zone Western Connecticut32FCA 9 Active DR (DRV) in load zone of Connecticut725.5% Distribution Losses+ 4Unavailable Active DR (25%)-19Active DR Modeled in Case57MaineFCA 9 Active DR (DRV) in dispatch zone Emera Maine27FCA 9 Active DR (DRV) in dispatch zone Maine97FCA 9 Active DR (DRV) in dispatch zone Portland ME17FCA 9 Active DR (DRV) in load zone of Maine1415.5% Distribution Losses+ 8Unavailable Active DR (25%)-37Active DR Modeled in Case111New HampshireFCA 9 Active DR (DRV) in dispatch zone New Hampshire13FCA 9 Active DR (DRV) in dispatch zone Seacoast2FCA 9 Active DR (DRV) in load zone of New Hampshire155.5% Distribution Losses+ 1Unavailable Active DR (25%)-4Active DR Modeled in Case12VermontFCA 9 Active DR (DRV) in dispatch zone Northwest Vermont25FCA 9 Active DR (DRV) in dispatch zone Vermont5FCA 9 Active DR (DRV) in load zone of Vermont305.5% Distribution Losses+ 2Unavailable Active DR (25%)-8Active DR Modeled in Case24Rhode IslandFCA 9 Active DR (DRV) in dispatch zone Rhode Island56FCA 9 Active DR (DRV) in load zone of Rhode Island565.5% Distribution Losses+ 3Unavailable Active DR (25%)-15Active DR Modeled in Case44NEMA/BOSTONFCA 9 Active DR (DRV) in dispatch zone Boston50FCA 9 Active DR (DRV) in dispatch zone North shore18FCA 9 Active DR (DRV) in load zone of NEMA/BOSTON685.5% Distribution Losses+ 4Unavailable Active DR (25%)-18Active DR Modeled in Case54SEMAFCA 9 Active DR (DRV) in dispatch zone Lower SEMA7FCA 9 Active DR (DRV) in dispatch zone SEMA41FCA 9 Active DR (DRV) in load zone of SEMA485.5% Distribution Losses+ 3Unavailable Active DR (25%)-13Active DR Modeled in Case38WCMAFCA 9 Active DR (DRV) in dispatch zone Central MA32FCA 9 Active DR (DRV) in dispatch zone Springfield8FCA 9 Active DR (DRV) in dispatch zone Western MA15FCA 9 Active DR(DRV) in load zone of WCMA555.5% Distribution Losses+ 3Unavailable Active DR (25%)-15Active DR Modeled in Case44New EnglandFCA 9 Active DR (DRV) in New England4845.5% Distribution Losses+ 27Unavailable Active DR (25%)-128Active DR Modeled in Case383Appendix C: PV generation ModelingAs part of the 2015 PV forecast the data on PV generation was divided into the following mutually exclusive categories:PV as a capacity resource in the Forward Capacity Market (FCM)Non-FCM Settlement only Resources (SOR) and Generators (per OP-14)Behind-the-Meter (BTM) PV Embedded in Load (BTMEL)BTM PV Not Embedded in Load (BTMNEL)Of the four categories, the BTMNEL is already embedded in the CELT forecast and hence will not be modeled explicitly in any studies. The remaining three categories need to be considered when accounting for PV generation in studies.The PV generation forecast only forecasts the PV values on a state-wide basis. However, within a state the PV does not grow uniformly, with some areas in the state having larger amounts of PV. To account for this locational variation of PV, the locational data of existing PV that is in-service as of the end of 2014 was utilized to obtain the percentage of PV that is in each dispatch zone. New England is divided into 19 dispatch zones and the percentage of PV in each dispatch zone as a percentage of total PV in the state is available. This percentage is assumed to stay constant for future years to allocate future PV to the dispatch zones. The percentage of existing solar in each dispatch zone as of the end of each year that is used as a part of the PV generation forecast is based on Distribution Owner interconnection data and the materials are located at: the PV generation data is organized by dispatch zone, the PV within the dispatch zone falls into three categories:Category 1 : Units greater than 5MW:Location data available Will be modeled as an individual generators13 MW of PV in MA are the only resources that fall into this categoryCategory 2 : Units greater than 1 MW and less than 5 MWLocation data available through the PPA notificationsNeeds to be modeled as injections at specific locations – Negative loads similar to DRCategory 3: Units below 1 MWNo location data availableNeeds to be modeled by spreading the MWs across the dispatch zone – Negative loads similar to DR and spread across the load zone/dispatch zone like DR is spreadFor PV in categories 2 and 3 the PV will be modeled as negative loads at the buses. The power factor of the negative load will be set to match the power factor of the load at the bus. If no load is present at the bus then a unity power factor will be assumed.Based on a review of historic PV outputs, ISO-NE Transmission Planning has determined a 26% availability factor is applied to all categories of modeled PV generation in transmission planning studies. The following table shows the PV generation modeled in the peak load case. Note that since the PV generation data is nameplate AC, a 5.5% gross up was performed to the PV generation forecast to account for avoided distribution losses.Table 17: PV generation Modeled in the Peak Load CasePV generation Modeled as Negative LoadsStudy YearLoad Zone2021 PeakConnecticutPV generation (nameplate) in dispatch zone Eastern Connecticut71PV generation (nameplate) in dispatch zone Northern Connecticut78PV generation (nameplate) in dispatch zone Norwalk-Stamford29PV generation (nameplate) in dispatch zone Western Connecticut178A - Total PV generation (nameplate) in load zone of Connecticut356B - 5.5% Reduction in Distribution Losses20C - Unavailable PV generation (A+B)*(1-26%)278PV generation Modeled in Case as Negative Loads(A+B-C)98MainePV generation (nameplate) in dispatch zone Emera Maine2PV generation (nameplate) in dispatch zone Maine4PV generation (nameplate) in dispatch zone Portland ME2A - Total PV generation (nameplate) in load zone of Maine8B - 5.5% Reduction in Distribution Losses+ 0C - Unavailable PV generation (A+B)*(1-26%)-6PV generation Modeled in Case as Negative Loads(A+B-C)2New HampshirePV generation (nameplate) in dispatch zone New Hampshire23PV generation (nameplate) in dispatch zone Seacoast3A - Total PV generation (nameplate) in load zone of New Hampshire26B - 5.5% Reduction in Distribution Losses+ 1C - Unavailable PV generation (A+B)*(1-26%)-20PV generation Modeled in Case as Negative Loads(A+B-C)7VermontPV generation (nameplate) in dispatch zone Northwest Vermont115PV generation (nameplate) in dispatch zone Vermont62A - Total PV generation (nameplate) in load zone of Vermont177B - 5.5% Reduction in Distribution Losses+ 10C - Unavailable PV generation (A+B)*(1-26%)-138PV generation Modeled in Case as Negative Loads(A+B-C)48Rhode IslandPV generation (nameplate) in dispatch zone Rhode Island144A - Total PV generation (nameplate) in load zone of Rhode Island144B - 5.5% Reduction in Distribution Losses+ 8C - Unavailable PV generation (A+B)*(1-26%)-113PV generation Modeled in Case as Negative Loads(A+B-C)40NEMA/BOSTONPV generation (nameplate) in dispatch zone Boston101PV generation (nameplate) in dispatch zone North shore53A - Total PV generation (nameplate) in load zone of NEMA/BOSTON154B - 5.5% Reduction in Distribution Losses+ 8C - Unavailable PV generation (A+B)*(1-26%)-120PV generation Modeled in Case as Negative Loads(A+B-C)42SEMAPV generation (nameplate) in dispatch zone Lower SEMA175PV generation (nameplate) in dispatch zone SEMA202A - Total PV generation (nameplate) in load zone of SEMA 377B - 5.5% Reduction in Distribution Losses+ 21C - Unavailable PV generation (A+B)*(1-26%)-294PV generation Modeled in Case as Negative Loads(A+B-C)103WCMAPV generation (nameplate) in dispatch zone Central MA171PV generation (nameplate) in dispatch zone Springfield59PV generation (nameplate) in dispatch zone Western MA221A - Total PV generation (nameplate) in load zone of WCMA451B - 5.5% Reduction in Distribution Losses+ 25C - Unavailable PV generation (A+B)*(1-26%)-352PV generation Modeled in Case as Negative Loads(A+B-C)124New EnglandA - PV generation (nameplate) in New England1694B - 5.5% Reduction in Distribution Losses+ 93C - Unavailable PV generation (A+B)*(1-26%)-1322PV generation Modeled in Case as Negative Loads(A+B-C)465Appendix D: General Comments on Stability CasesPer the latest approved models, Kingdom Wind (KCW), Bull Hill Wind, and Pisgah Wind use VESTASWT_7_4_4 models. KCW, however, has a customized version of this model owing to its configuration. It appears that KCW cannot be dispatched with Bull Hill Wind or Pisgah Wind online in the same case. Transmission planning is working with the developer to resolve this issue. In the interim, though, these three plants have been included and compiled in the case, with Bull Hill Wind and Pisgah Wind turned off. KCW unit can be dispatched in this stability package. Some cases may issue the following message during initialization. “BUS 103140 [BULLHLL_GEN 0.6500] MC 1 USER DEFINED MACHINE MODEL "VWEX1" NOT ACCESSIBLE” You may also see the following suspect initial condition. I DSTATE(I) STATE(I) MODEL STATE BUS# X-- NAME --X BASKV ID 117597 0.10000E+00 -0.19817E-05 VWGOV K+3 103140 BULLHLL_GEN 0.6500 1 Please note that that these messages have no bearing on the simulations. PSS/E v32.2 has been used to create the stability cases. Please make sure that the bus dimension is set to 150,000 (BUSES 150000) before running a simulation. Steps for running the simulation: 1. Open PSS/E saved case from the received package. 2. After making any study specific dispatch changes run the ‘convert.idv’ automation file and save the converted case. 3. Load the converted case and the snapshot provided to run simulation. Note 1: The files “oel_uel.dll” and “URRTGT_rev32.dll” are user-written library models for certain units, and are not part of the main dsusr.dll. These files must be present in the same directory as the “dsusr.dll” file, and must be loaded into PSS/E during stability simulations. One of the following methods can be used to load these additional dll files:Go to PSS/E Tools – Load Model Library, and load these two libraries separately. The main dll file “dsusr.dll” will be loaded automatically as long as it is in the same directoryAlternatively, you can use the following command BAT_ADDMODELLIBRARY in the PSAS file to load these dll files separately. Please remember to update the directory locations where the dll files are located. For example: BAT_ADDMODELLIBRARY,'C:\xxx\xxx\xxx\xxx\xxx.dll'Note 2: Initialization errors can occur if the steady state reactive output of the Marcy, Fraser, or Leeds SVCs in NYISO is too high. Moreover, some minor machine rotor angle deviations were observed during the ring down test of the stability package when the Coopers Mills STATCOM output is not near 0 MVAR or when the Detroit SVC output is not above 0 MVAR. “NYSVC_plus.py” reduces the 3 NYISO and Coopers Mills svc/facts reactive output to zero by changing their voltage schedules equal to what is in the open case. This python file sets the output of the Detroit SVC to 10 MVAR by modifying the voltage schedule. This python file should be run after making significant dispatch changes, especially of NY and ME machines. Note that this python file solves the case with the following solution options: transformers regulating, switched shunts regulating, phase angle regulators regulating, and area interchange enabled. If different solution options are used then the python file will need to be modified accordingly.Note 3: The file “V82BCPMX1.DAT,” must be present in the same directory as the DSUSR.dll, otherwise PSS/E dynamics crashes. This file links to certain Vestas wind farm models. Note 4: The file “CP_GW90.txt” must be included in same directory as all other data in order for Georgia Mountain Wind user model to function correctly.Note 5: With the inclusion of the 326 SPS thermal mode relay model, to use this model when simulating with faults that include opening Line 326, a small value of shunt MVA (1-j1) is needed on either the Scobie Pond or Lawrence Road 345-kV buses and the Scobie Pond-Lawrence Road section of Line 326 needs to stay in-service. This allows a negligible MVA to still be flowing on Scobie Pond-Lawrence Road section of Line 326, otherwise the model will trip the assigned branches (generator GSUs), even though the thermal set point may not have been reached. Opening of the Scobie Pond-Lawrence Road section of Line 326, which is being monitored by the SPS model, causes the assigned branches to be tripped (and thereby tripping the units associated with this SPS) after the associated time delay; this is inherent in the PSSE relay model. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download