The URL can be used to link to this page

Home My WebLink AboutBradley Lake PMC Meeting Tuesday 06-20-2000Notification of Bradley Lake Project Management Committee Regular Meeting A regular meeting is scheduled to be held on Tuesday, June 20, 2000, at 10:00 a.m. at the Chugach Electric Association’s Training Room located at 5601 Minnesota Drive, Anchorage, Alaska. Packets are scheduled to be sent out on Friday, June 9, 2000. Please call Shauna at (907) 269-3028 with questions. h:\all\sdean\bradley\mtg notification fax cover.doc PLEASE DELIVER TO THE FOLLOWING: Eugene Bjornstad Norm Story Wayne Carmony Hank Nikkels George Kitchens Dave Calvert Ron Saxton Chugach Electric Association Homer Electric Association Matanuska Electric Association Anchorage Municipal Light & Power Golden Valley Electric Association City of Seward Ater Wynne LLP (907) 562-6994 (907) 235-3323 (907) 745-9368 (907) 263-5204 (907) 458-5951 (907) 224-4085 (503) 226-0079 RK KKK R111 C COCO CCK BROADCAST DATE START — RECEIVER x x x X x X X X x MAY-26 10:49 AM 95626994 x 10:50 AM 919072353323 k 10:51 AM 919077459368 k 10:52 AM 92635204 x 10:54 AM 919074585951 x 10:55 AM 919072244085 k 10:56 AM 9150322600793 X x x x X TRANSACTION REPORT TX TIME PAGES TYPE 46” 3 SEND 41" 3 ‘SEND 44” 3 SEND 1'50”" 3. SEND 42" 3 SEND 45” 3° SEND 47" 3s SEND TOTAL P, O1 x X MAY-26-00 FRI 10:57 AM x x x X NOTE Mt DP x X ( M) OK 001 x (M) OK 001 x ( M) OK 001 k ( M) OK 001 x (M) OK 001 x (M) OK 001 x (M) OK 001 x X X 6M 15S PAGES: 21 k X X 0010001000001 1110000111100 00000 OOOO COCO C CCC CCC COCO NOOO C OOOO PLEASE DELIVER TO THE FOLLOWING: Eugene Bjornstad Norm Story Wayne Carmony Hank Nikkels George Kitchens Dave Calvert Ron Saxton Chugach Electric Association (907) 562-6994 Homer Electric Association (907) 235-3323 Matanuska Electric Association (907) 745-9368 Anchorage Municipal Light & Power (907) 263-5204 Golden Valley Electric Association (907) 458-5951 City of Seward Ater Wynne LLP (907) 224-4085 (503) 226-0079 us ALASKA INDUSTRIAL DEVELOPMENT AND EXPORT AUTHORITY {= ALASKA @™ =ENERGY AUTHORITY 480 WEST TUDOR ANCHORAGE, ALASKA 99503 907 / 561-8050 FAX 907 /561-8998 Bradley Lake Instability Study Final Report June 9, 2000 Prepared by: Brian Hickey Randy Johnson Larry Hembree Page 1 of 15 Executive Summary The Bradley Lake Instability Project Team has completed its analysis and identification of the instability and control response problems at the Bradley Lake Project. We have identified the source of both problems and are recommending that the O & D and the PMC fund proof of concept designs on a number of proposed modifications to the plants control systems. These concepts are explained in the following report. The project team has determined that the Bradley Plant is the source of the 560-750ms oscillations identified in the 1997 Kenai disturbance review. We have determined that the oscillations are the result of small signal instability in the Bradley control system. This instability is caused by time delays, gain mismatches, system non-linearity’s, and poor governor control algorithm design. The team has determined that the control response problems are a result of control algorithms that focus on minimizing water usage over performance. These algorithms were not modeled in the studies that were used to predict the plants response and develop the final operating guidelines. The project team believes it is technically possible to eliminate these shortfalls. And, that plant performance can be improved by replacing the electronic governor hardware and software, mitigating turbine pit turbulence, adding dual rate servo motors, and adding external plant intelligence to allow real-time optimization of gains. The team has completed preliminary cost benefits analysis on the economics of three alternatives: 1. Do nothing, limit plant output to 0 MW spinning reserve and 60 MW capacity. No greater than 30 MW per unit. 2. Fix instability and accept the plants current non-responsive spinning reserve contribution. 3. Fix the instability and improve control response. Assuming a worst case cost for alternative three of $5 million dollars, the fuel and O & M benefits alone provide a seven year payback. Seven years is well short of the remaining life of the plant. The project team recommends that we proceed with alternative three. The team recommends that we proceed with funding for proof of concept designs and request quotations on these designs. Upon receipt of these quotations additional cost benefit analysis will be performed and construction funding will be requested for those options with positive cost benefit ratios. The project team request funding in the amount of $796,200 to complete project design and enhancement tasks identified in appendix A of the report. Page 2 of 15 Upon completion of the project enhancements, underfrequency loadshed and final operating guides, and tuning of the existing Static Var Compensator Power Oscillation Damper will be required. Funding for these studies is not included in this request. Page 3 of 15 Background In January 1996 at the request of Chugach Electric, the Bradley Lake dispatcher, the Bradley Lake Operations & Dispatch subcommittee (Bradley O & D) recommended and the Bradley Lake Project Management Committee (BPMC) subsequently authorized an investigation into the cause of numerous electrical instability events and control response problems that had been occurring since the commissioning of the Bradley Lake project. A project team consisting of Larry Hembree (AMLP), Randy Johnson (AML&P), Brian Hickey (Chugach), was assembled. The team was supported by Andi Hamilton, Budget and Project Coordinator (Chugach); Don Stead, Bradley Project Manager(Homer); Stan Sieczkowski,AEA representative and numerous consultants including Electric Power Systems (EPS), Woodward Governor Co. Inc., Stone & Webster Inc., Power Technologies, Inc. Initially the investigation consisted of compiling a comprehensive review of Bradley Project documentation related to technical design and commissioning notes and a review and analysis of all electronic power system data related to system disturbances since Bradley commissioning (1993). The systematic review Bradley documentation validated the two premises of the investigation. First, that the measured control response of the plant is significantly poorer than that predicted in the design studies and second that electrical instability had been identified documented and not resolved during commissioning. The review of power system data involved analysis of 90 disturbance events and from them identified approximately 30 oscillations. This review empirically quantified the period of the oscillations as 560-750 ms and validated that they did in fact exist and that Bradley was always a significant participant if not driver of them. These results led to the development of the action plan for correction of the Bradley Lake Project instability and control problems. The action plan was submitted to the O & D and subsequently authorized and partially funded by the BPMC in January of 1997. The action plan detailed a series evaluations studies, and modeling projects intended to answer several key questions raised in validating the suspected causes of the electrical instability and control response problem. These questions are: Regarding Electrical Instability: i From a power system perspective, does the Bradley Project initiate instability or does the Kenai Power System or some other external sources (Static VAR Compensator, CEA-SCADA etc.) excite the Bradley Project into instability? 2s If Bradley is the source of instability, what control system or systems cause, or contribute to the cause, of the instability? Page 4 of 15 35 Why doesn’t our PSS/E model replicate these oscillations? And, can they be made to do so? Regarding Control Response: Ie Why is Bradley’s control response significantly slower than predicted in the 1990 Load Acceptance study, the 1992 Dispatch Operations Summary, and the 1993 Final Operating study? What is the plants actual maximum response rate in needle mode? 2. What control components contribute to the under performance of the plant? Be Can control response be improved, if so how and what is the cost? Regarding system operation: The action plan details studies to be performed at the conclusion of the project to provide accurate operating guidelines to system operators; and to modify existing power system protection and reliability schemes to optimize plant performance and system reliability. One of the key challenges in answering the questions posed was the apparent inconsistency between the 1990 Bradley Lake PSS/E Tunnel-Hydro-Turbine-Generator Model developed by PTI and the actual Bradley Lake Tunnel-Hydro-Turbine-Generator control system. The project team engaged EPS to review and document the actual plant control system, to evaluate the PTI model, and to develop and bench mark a new more accurate Bradley Governor model. This project was completed in January of 1998. The model was able to reproduce the instability with the observed modes of oscillation; however, the project team determined that additional tuning was required to accurately predict instability and to better understand governor control quantities. Additionally, as a result of the EPS modeling evaluation several revisions were made to the railbelt PSS/E model to more accurately reflect actual power system devices external to Bradley. Another key challenge to understanding the Bradley electrical instability phenomenon was the time synchronized acquisition of a relatively large number of data points. The railbelt utilities did not possess an instrument of sufficient bandwidth to gather this data and were unable to locate one commercially. Therefore, the project team designed and built the device utilizing “off the shelf’ components and commercially available software. The device, known as “The Instrumentation System”(IS) samples 72 channels of data 2000 times per second, with less than .006milliseconds (ms) of phase shift between channel 1 and channel 72. The system is used to record power quantities, internal governor quantities and other physical parameters of the plant. The device was installed, commissioned and tested at Bradley in May of 1999. Page 5 of 15 Concurrently, the project team developed a detailed Bradley Turbine-Governor Test Plan. This plan was reviewed and approved by Woodward Governor , Electric Power Systems and Stone and Webster. The test plan was structured to answer the following questions: 1. Ds Is the Bradley Instability a classic case of small signal instability? If 1 is true, what is the source of the small signal perturbations: monaogep Governor/Plant Electronics? Tunnel Pressure Waves? Turbine Pit Turbulence? Deflector Vibration? Governor Hydraulic System Transients? Kenai Power System Noise? What, if any, are the contributions of the following components to instability and if contributions are made, what causes the perturbation or time delays? sae rp Ss Eda rh Speed Transducer Time Delay? Deflector PID Gain Settings? Governor Processing Time Delay? Hydraulic Servomotor Time Delay? Deflector Speed Variations Due to Hydraulic Pressure Variations? Deflector Position Telemetry Error? Exciter Interaction? Power System Stabilizer Interaction? SVC Power Oscillation Damper Interaction? Deflector position error? Can a more accurate deflector verses power curve be derived? Does the new EPS tunnel model more accurately reflect the physical tunnel? Can the instability be reproduced routinely? What is the maximum rate at which needles can be pulsed full open to full closed and vise versa? In May of 1999 testing was performed on the Bradley Plant and data was collected utilizing the IS system. Woodward Governor was on site to assist in connecting the IS system to the governor and to reconfigure governor algorithms to export real time data from the 501 controller directly to IS data collection channels. Stone and Webster and EPS were on-site to provide assistance on Plant and power system technical issues. Page 6 of 15 Twenty-five gigabytes of raw data were collected over the five days of testing. Over the past year the project team has developed analysis algorithms, correlated and collated the data, and has arrived at answers to each of the questions the test plan was developed to answer. In addition, the team benchmarked all stability-related components of the PSS/E model by isolating & modeling each governor sub-component in MATLAB (a GUI Control Algorithm modeling program). Actual system data, gathered from the machine tests has been streamed into the model sub-component and the output of the sub-component subtracted from the actual output of that sub-component during testing. This function represents the error for a given sub-component, for that input function. Numerous representative input functions were run through each sub-component and the worst case errors were identified. Current Status of the Project We have pinpointed the nature and cause of both the instability and control response problems at the plant. In addition, we have identified what we believe are methods to resolve these problems. We have identified three alternatives to mitigate the problems. And have evaluated the Railbelt wide fuel, and O&M costs and benefits of these alternatives. Finally, we have prepared our recommendation as to where we should go from here. The following is a list of answers to questions we set about to answer in this investigation: Question From a power system perspective does Bradley initiate the instability, or does the Kenai Power System or some other external entity initiate the instability? Answer The Bradley Instability is a classic case of small signal instability. The Bradley Control System causes the instability. The system is perturbed by Turbine Pit turbulence aided at times by Kenai power system load variations. Question If Bradley is the source of instability what control system components contribute to or cause the instability? Answer Instability is a complex phenomenon, however in short, it is caused by positive feed back & excessive gain in a control system. In the Bradley Plant the positive feed back is caused by time delays between the time that the rotor speed is captured by the turbine shaft speed transducer and the time that this Page 7 of 15 speed, compared to desired speed, is translated into a new physical deflector position. The gain settings within the controller aggravate this condition. The overall delay is approximately 370 milliseconds. To put this in context, a controller delayed by 370 milliseconds is controlling oscillations with periods ranging from 560 to 750 milliseconds. Components that contribute to this delay are: ie Speed transducer 200ms 2: Governor Processing 010ms 35 Hydraulic Servos 160ms The non-linearity of the deflector vs power curve and the speed of the deflectors (1.5 seconds from 0 deflection to full deflection) aggravate this problem. The problem is further aggravated by the fact that the control scheme for the deflector contains no droop nor is there any needle speed reference (NSR) feedback when deflectors are in stream and the unit is in needle mode (the typical overspeed scenario). Zero droop causes the machine to operate isochronously and without NSR feedback the deflectors remain in stream indefinitely or until Automation Generation Control (AGC) or the dispatcher schedules a new, lower, set point. Finally, so long as the deflectors are in stream, Turbine Pit turbulence creates random power system noise. This noise is approximately 4-6 MW in magnitude per generator and is sufficient to perturb the unstable reaction. In summary the problem is a result of control response time delay, non-optimal gains and is aggravated by a lack of droop, a lack of power system feedback, and small signal noise generated by turbulence within the turbine pit and caused by the deflectors. Question Why doesn’t the PTI PSSE Model allow us to reproduce the oscillations? Answer The original PTI Model did not accurately represent the Woodward 501 Governor or tunnel. The EPS Model accurately represents the 501 Governor and tunnel. Tuning performed by the project team, following machine testing has allowed us to recreate oscillations and is as accurate a digital representation of an analog system as is practical. Control Response Problem Bradley Lake was designed with 72 seconds (90s nominal) needle opening and closing times. This means that the needles are capable of fully loading or unloading the unit in 72 seconds. Based on an analysis performed in 1990 by Stone and Webster/PTI, the project was designed to provide 27 MW of responsive spinning reserve when operated in conjunction with the Railbelt combustion turbines. Bradley was designed to provide 27 MW of spinning reserve in 15-18 seconds. It was intended that the Railbelt combustion turbines would cover the Bradley spin response deficiency by operating in peak mode (10% above thermal limit) for this 15-18 second period. Page 8 of 15 The question was posed: Why is Bradley’s control response significantly slower than predicted in the 1990 Load Acceptance study and final Bradley Operating Studies? What is Bradley’s maximum response rate? Answering the second question first, for the current governor the maximum obtainable open/close rate via AGC is approximately 8 minutes. Under some conditions the opening rate can be as slow as 25 minutes. Why the slow response? The answer lies in the needle scheduling logic of the Woodward 501 Governor. The governor is designed to minimize water usage and therefore schedules the six Bradley needles in 3 pairs. The needles transition from 2 needles to 4 @ 18 MW & from 4 needles to 6 @ 36 MW. At each transition point needle equalization logic takes control. For example, at the 2-4-needle transition the two fully open needles transition closed, while the 2 fully closed needles transition open, until the needle opening for all needles is equal. No additional power output is generated during the equalization period. The slow control response is exacerbated by the inability of speed error to schedule needle transitions. Needle transitions can only be scheduled by needle speed reference changes initiated by AGC or the plant operator. The gist of this is that, the capacity that the unit can provide in droop response is limited to the difference between its energy output and the next needle transition. For example, a single unit loaded at 16 MW can provide 2 MW ( 18MW-16MW) of response without AGC or operator intervention. In summary the control response problem is a function if the Woodward 501 control algorithms. These control algorithms are the first barrier to improved control response. The next barrier is the 90 sec needles and closely interrelated to that is the tunnel hydraulics. Control response can be improved; however, not with the current version governor electronics. Alternatives The project team has settled on three potential alternatives to provide the O&D and subsequently the BPMC with cost benefit evaluation information. The alternatives are: IF Do nothing to the plant and adjust operating guidelines to minimize the potential of instability This option provides no spinning reserve individual unit out limits of 30 MW with the total plant out put limited to 60 MW. 2: Replace Governor electronics to reduce time delays. If possible, dampen turbine PIT noise, add external intelligence to optimize gains based on power system conditions. Successful implementation of these concepts Page 9 of 15 can virtually eliminate instability and potentially provide the current 27 MW of non-responsive spin. 3. In addition to number two, install dual-rate hydraulic servomotors and increased needle speeds. Implement a dual mode algorithm, “efficiency” verses “system support”. Efficiency mode would maintain needle equalization and System Support mode would trip to six needles open as fast as possible. The economic analysis of these alternatives was performed independently by John Cooley of Chugach and Bob Price of AML&P utilizing each companies respective production costing model. Each analyst made his own assumptions as to how the system would operate under each alternative. In this way we were able to demonstrate qualitatively the magnitude of the variation in cost/benefit depending on system operating assumptions. System operating assumptions are assumptions such as: When and if Healy Clean Coal is returned to service, how much silos GVEA arms, or when Nikiski Cogen becomes base loaded etc. These costs and benefits are strictly operating costs, fuel and O & M. No value was placed on spilled water, increased reliability, reduction in wear & tear or potential catastrophic damage resulting from instability. Alternative one as compared to current operating practice costs the interconnection between $689,000 & $1,200,000 annually. Alternative two, utility costs remain the same as today. Alternative three shows an annual savings of $1,400,000 to $2,000,000 over alternative one and $711,000 - $800,000 over current operating practice. We have estimated that a complete rework of the Governor electronics, hydraulics, and needle orifices would not cost no more than 5 million dollars. Assuming the worst case benefit of $711,000 annually and 5 million dollar investment the pay back period on this investment would be approximately 7 years, well short of the remaining life of the Bradley plant. Recommendations The Project team recommends completing proof of concept designs and studies for the enhancements required to achieve alternative three. We propose that upon completion of these designs, we request firm quotations from vendors on each enhancement. At that time, we will perform a more detailed cost benefit analysis of each alternative. This will allow the BPMC to selectively fund enhancements in order to maximize benefits. We propose the following process to insure that the proof of concept designs are valid. Page 10 of 15 e Where appropriate models will be required for each design. These models will be critiqued by the team for validity and practicality. Each model will be tested against the power system utilizing MATLAB and PSS/E. e Where appropriate, designs will be translated into functional specifications with typical design drawings attached. e Requests for quotation will be issued to appropriate vendors. e RFQs will require the vendor to respond with control design, vendor model and quotation. e These vendor models will be critiqued and tested against the power system. e Upon completion of testing and cost benefit analysis the team will return to the BPMC to authorize funding for construction. e All final designs will be provided with an appropriate test plan. These test plans will provide measurable performance criteria to which financial penalties will be attached. Penalties will be implemented if enhancements fail to meet specification. This process provides the utilities with the opportunity to review, comment on, and test designs prior to construction. If modeling provides unacceptable results, or the cost- benefit ratio of the enhancement is inadequate this process provides an opportunity for the BPMC to cancel construction funding. The Project Team recommends the Operations and Dispatch Subcommittee, and subsequently the Bradley Lake Project Management Committee, authorize funding in the amount of $796,200 to complete proof of concept design and testing on the following enhancements to the Bradley lake Project. Further detail on these enhancements can be found in Appendix A. e Design Fiberoptic Shaft Speed Transducer $ 80,000 e Governor control algorithm design $105,000 e Develop functional specification of design $ 25,000 e Design external intelligence and SCADA interface $ 50,400 e Identify the nature of the Turbine Pit turbulence $ 30,000 e Design Turbine Pit turbulence mitigation $120,000 e Enhance and harden Instrumentation System $ 40,200 e Review and testing of vendor supplied models $ 40,000 e Design Power Oscillation Damper for Soldotna-Bradley Lines $ 70,000 e Evaluate replacement of Exciter and PSS with integrated model $ 15,000 e Railbelt Peak Mode droop coordination design $ 20,000 e Design dual rate Hydraulic Servo motor system $125,000 e Design and implementation of interim instability mitigation $ 33,600 e Design condense to generate logic and SCADA interface $ 33,600 e Develop an project implementation and construction plan $ 8,400 Total Funding Requirement: $796,200 Page 11 of 15 Interim Mitigation In the interim between today and completion of the proposed enhancement the following measures will be taken to minimize the likelihood of additional unstable oscillations. Short Term 1. Deflector derivative gains set to zero in deflector. 2. Unit operated in inflexible AGC mode. 3. Alarm generated when deflector is suspected of being in stream. 4. Dispatchers 1* priory response to alarm, is to lower unit output and system frequency if necessary at the expense of load shed. D5 Man the Bradley Lake Control Room whenever units are loaded above 30 MW. 6. Develop Interim Dispatch Operating Guide Long Term 1. Install unit trip hot button in Dispatch 2. Automate frequency reduction algorithm 3. Telemeter jet edge trigger to Dispatch 4. Optimize Gains to minimize instability Page 12 of 15 Appendix A Rates used in these estimates are $125/hour for consultants and $105/hour for utility employees. Shaft Speed Transducer Design fiber-optic shaft speed transducer $ 80,000 Consultant Duration 2 months $125/hr Governor Control Algorithm Design $105,000 Performed primarily by 50% utility personnel 50% consultant Includes: e Efficiency and support needle mode. e Deflector control with droop and speed reference feedback. e Elimination of MSR chatter. e Provisions for real time gain optimization. Duration — 6 months $105/hr Develop a Functional Specification $ 25,000 Develop a functional specification from the Governor control algorithm design. Consultant w/Utility Review Duration 5 weeks 200 MH @ 125/m Design of External Intelligence and SCADA Interface $ 50,400 Design of a PLC or microprocessor based system that monitors Power System quantities, unit status, import/export values, unit Loading, system frequency etc. and can modify governor gains And issue instructions to master station equipment. Duration 3 months $105/hr Identify the Source of Turbine PIT Turbulence 5 days on site testing 2000/day 10,000 $ 20,000 Report 80 hrs @ 125/hr 10,000 $ 10,000 $ 30,000 Duration 3 months Design Turbine Pit Mitigation Devices $120,000 6 months @ $125/hr Page 13 of 15 Enhance and Harden Instrumentation System Upgrade I/O & triggering capability $ 40,200 COM Port & Software 15,000 Materials 25,200 Labor Duration 6 weeks Review and testing of Manufacturer Supplied Models $ 40,000 Review and testing of the vendor models supplied In response to the RFQ Duration 2 months 125/hr Design of POD Installation of Soldotna Bradley Lines $ 70,000 Quote from ABB — $10,000 Duration 3 months Integrated Exciter Power System Stabilizer Design $ 15,000 Evaluate replacement of Fuji Exciter/PTI PSS with Integrated Exciter/PSS Duration 3 weeks_$125/hr Railbelt Peak Mode Droop Coordination Design $ 20,000 Quantify peak-mode capacity contribution and duration for all Railbelt combustion turbines and design a coordinated Droop setting plan to maximize this value. Duration 1 month $125/ hr Design a Dual Rate Servo Motor System $125,000 Design dual rate hydraulic servo motor system for Needle operation. The system will provide faster speeds on opening than on closing. Duration 6 months _$125/hr Design and Implementation of Interim Instability Mitigation $ 33,600 Design and implement jet edge trigger to SCADA and SCADA response. Implement a Dispatch unit trip” hot button”. Optimize gains and develop interim operating guideline for the power system dispatchers . 2 month @ $10S5hr Page 14 of 15 Design Condense to Generate Logic $ 33,600 Design logic to issue a setpoint to the governor when tripped from condense to generate an under frequency event. 2 month @ $105/hr Develop Project Implementation and Construction Plan $ 8,400 2 weeks @ $105/hr Page 15 of 15 Bradley Lake Proof of Concept/Design & Studies ID__|Task Name Duration Start Finish Jun Jul | Aug | Sep | Oct | Nov | Dec - Feb | Mar | Apr | May | Jun Jul | Aug | Sep | Oct | Nov 1 Bradley Proof of Concept/Design & Studies 367 days Thu 07/06/00 Fri 11/30/01 | : rn eat 2 Design & Implementation of Interim Instability Mitigati 8wks Thu 07/06/00 Wed 08/30/00 Project Team,Dispatchers 3 Enhance/Harden Instrumentation System 6wks Thu 08/31/00 Wed 10/11/00 R. Johnson a 4 Governor Control Algorithm Design 25 wks = Thu 10/12/00 Wed 04/04/01 Project Team +, 5 Develop Functional Specification Swks Thu 04/05/01 Wed 05/09/01 Consultant + __ 6 Identify Source of Turbine Pit Mitigation Devices 12 wks Fri 09/01/00 Thu 11/23/00 Consultant 7 Design Tumine Pit Mitigation Devices 24wks = Fri 11/24/00 Thu 05/10/01 Consultant Selene “e | Railbelt Peak Mode Droop Coordination Design 4wks Tue 08/01/00 Mon 08/28/00 Machine Rating Subcommittee,Consultant 9 Integrated Exciter Power Sys Stabilizer Design 3wks Wed 11/01/00 Tue 11/21/00 Consultant 10 Design a Dual Rate Servo Motor System 25 wks Mon 06/11/01 Fri 11/30/01 Consultant 11 Design External Intelligence & SCADA Interface 12wks Tue 05/01/01 Mon 07/23/01 Consultant - SCADA Analyst 12 Design Condense to Generate Logic 8wks Tue 07/24/01 Mon 09/17/01 Consultant - SCADA Analyst +___ 13 Design of POD Installation of Soldotna Bradley Lines 12wks Thu 03/01/01 Wed 05/23/01 ABB 14 Develop Project Implementation & Const Plan 2wks Thu 11/01/01 Wed 11/14/01 Consultant 15 Design Fiber Optic Shaft Speed Transducer 8wks Wed 11/15/00 Tue 01/09/01 Consultant 16 Review & Testing of Manufacturer Supplied Models 367 days Thu 07/06/00 Fri 11/30/01 Consultant,Project Team ] Task Milestone r Rolled Up Split occ vee vas, External Tasks a Project: Bradley Lake Proof of Conce | spit Summary ey rolled Up Milestone © Project Summary TD Progress Rolled Up Task Rolled Up Progress Page 1 Shauna Dean From: Shauna Dean Sent: Friday, June 09, 2000 3:13 PM To: Cindy Crago Subject: Ad for Journal/Web Page Importance: High ww Admin Journal Ad.doc Here is an ad for the Admin Journal/Web Page. Title: Bradley Lake Project Management Committee Meeting Archive Date: June 19, 2000 Category: Committee Meeting Location: Anchorage, Alaska r ALASKA INDUSTRIAL DEVELOPMENT e_ AND EXPORT AUTHORITY /= ALASKA @@E™ =ENERGY AUTHORITY 480 WEST TUDOR ANCHORAGE, ALASKA 99503 907 / 269-3000 FAX 907 / 269-3044 June 9, 2000 FAXED: 06/09/00 2 PAGES Anchorage Daily News 1001 Northway Drive Anchorage, Alaska 99508 ATTENTION: Ms. Eva Alexie Legal Classified Subject: Public Notice Account # ALASO709 Enclosed is an advertisement for the Alaska Energy Authority, which is to be published under “PUBLIC NOTICE” in the Legal Classified section in column format. This notice needs to be published for one day: Monday, June 12, 2000. After the publication date, please provide an Affidavit of Publication. If you have any questions, please do not hesitate to call me. Thank you, Dea Shauna Dean Admin. Assistant Attachment 0000000001001 0000000000100 1001000000000 0000000000000 COC COCO X XK XK XK X X X X X X Xx X X Xx TRANSACTION REPORT yeoe-00 Rt osm SEND (M) DATE START — RECEIVER TX TIME PAGES TYPE NOTE Mt DP JUN-09 03:13 PM 92796170 35" 2 SEND (M) OK 124 TTT TOTAL 35S PAGES: 2 OOOO OOOO OOOO OOO OOOO OOOO OOO OOOO OOOO OOOO OOOO OOO OOOO OOOO OOOO OOO COICO OKO CLOAK ALASKA INDUSTRIAL DEVELOPMENT > AND EXPORT AUTHORITY a> ALASKA : @m™ =ENERCY AUTHORITY 480 WEST TUDOR ANCHORAGE, ALASKA 99503 907 / 269-3000 FAX 907 / 269-3044 June 9, 2000 FAXED: 06/09/00 2 PAGES — Anchorage Daily News 1001 Northway Drive Anchorage, Alaska 99508 ATTENTION: Ms. Eva Alexie Legal Classified Subject: Public Notice Account # ALAS0709 Enclosed is an advertisement for the Alaska Energy Authority, which is to be published under “PUBLIC NOTICE” in the Legal Classified section in column format. This notice needs to be published for one day: Monday, June 12, 2000. X X X X X X x Xx X X X x X X is ALASKA INDUSTRIAL DEVELOPMENT AND EXPORT AUTHORITY = ALASKA @@E™ ENERGY AUTHORITY 480 WEST TUDOR ANCHORAGE, ALASKA 99503 907 / 561-8050 FAX 907 /561-8998 Shauna Dean From: Dianne Hillemeyer [Dianne_Hillemeyer@chugachelectric.com] Sent: Thursday, May 25, 2000 3:05 PM To: sdean@aidea.org Subject: Re: Hi, Shauna - | guess Gene talked to most of the guys and they were okay with 6/20 (and now Brian will be here). So would you like to notice the meeting for 6/20 at 10:00 a.m. in Chugach's training room? Thanks much Ay = ZS BRADLEY LAKE PROJECT MANAGEMENT COMMITTEE MEETING TELECONFERENCE AGENDA Tuesday, March 20, 2001 — 10:00 a.m. (Via electronic media at AIDEA/AEA — 813 W. Northern Lights Boulevard) CALL TO ORDER Bjornstad ROLL CALL (for Committee members) PUBLIC ROLL CALL (for all others present) PUBLIC COMMENT AGENDA COMMENTS APPROVAL OF MEETING MINUTES — June 20, 2000 ies ee B reste COMMITTEE COMMENTS A. Next Meeting Date Bjornstad ADJOURNMENT * * * * @ ALASKA INDUSTRIAL DEVELOPMENT = ATEN “mc Bcortaron" "= ALASKA = _ENERGY AUTHORITY —— *k * 813 WEST NORTHERN LIGHTS BLVD. ANCHORAGE, ALASKA 99503 907 / 269-3000 FAX 907 / 269-3044 Facsimile Transmittal re : TO: (2€ne Bjocnstad COMPANY: Chita Electric. Assoc¢at}om~ Fax #: (%0F)__ 562-6797 FROM: haw. Dea aa DATE: 9 O(- O| TIME: [12 on. Number of pages including cover page: Transmittal Contents: Comments: —~ C Wii Notice: This facsimile may contain confidential information that is being transmitted to and is intended only for the use of the recipient named above. Reading, disclosure, discussion, dissemination, distribution, or copying of this information by anyone other than the named recipient or his or her employees or agents is strictly prohibited. If you have received this facsimile in error, please immediately destroy it and notify us by telephone at (907) 269-3000. H:\ALL\ERYN\Facsimile Transmittal Cover Page.doc DARARAARRARADARADARA DADA DAARE BARA BARA REARRR ABABA RAR AAR RRR ARBRE RRA IRA RRR AIR RRR ROK A AOR ORK x P. O01 x k TRANSACTION REPORT x k _—_— JAN-03-01 WED 09:58 AM x K K x SEND (M) x x k x DATE START — RECEIVER TX TIME PAGES TYPE NOTE Mt DP x X ik x JAN-03 09:54 AM 95617683 3'40" 8 SEND ( M) OK 072 x sing x x x TOTAL 3M 40S PAGES: 8 k k x SOOOOOOOOOOOOOO OOOO OOOO OOOO OOOO OOOO OOOO OOOO OOOO OOOO OA OOOO OOO OOOO OOOO OO OOOO OO OOOO OOO KKK ALASKA INDUSTRIAL DEVELOPMENT > ¢ AND EXPORT AUTHORITY f= ALASKA MI = ENERGY AUTHORITY 813 WEST NORTHERN LIGHTS BLVD. ¢ ANCHORAGE, ALASKA 99503 © 907/269-3000 ¢ FAX 907/269-3044 TOLL FREE (ALASKA ONLY) 888 / 300-8534 Facsimile Transmittal TO: Xp. Koole 7 a COMPANY: = Lj Fax #: (FOP) SES £83 FROM: rawr Lea DATE: 73-07 TIME: 2:45 a.m. Number of pages including cover page: - e Transmittal Contents: STATE OF ALASICR, —/ roman comme DEPARTMENT OF COMMUNITY AND 339 WEST FOURTH AVE., SUITE 220 ANCHORAGE, ALASKA 99501-2341 REGIONAL AFFAIRS ane nied DIRECTOR'S FAX: (907) 269-4645 DIVISION OF ENERGY ENGINEERING FAX: (907) 269-4685 & printed on recycled paper b y C.D S$ -220- 00 “pb E a $e ae —— Coa pat iE == sea aoe DEA ai , geen yi ae a, ee ltr. lovbive — Kenoic, Cay YOncis Lorie (ie seemdem +) Pot Qerr-ok = = : - si Stead HeA 00 See acyl Laven tre, HEA Ace _f Dave ¢ me od ats Dens ea AeA "F50O milli ogee ar OLA: one Cyole op ies GIEA us fisxds of 00 ochions ae fine taney Be > LAE A Bria Hi Bob D Lap alt CEA —ldacted by, sath Jodo Blige folloEis Gok Vas it tue driven =a “Pod. me Sve Hak = Pe Suk te a epee BL opschon A“ Mam Stay Zaceer Dload Carrols” apa : @ bhi led . | =) Coxe thos ped. | “Yoase dap iia Ht Play Prana = faving syuee > Peading fete Hele Vth dy oD Comat Med Hey. fain Mace! at AdfienO, —#. BRADLEY LAKE PROJECT MANAGEMENT COMMITTEE MEETING MINUTES Via Electronic Media @ Chugach Electric Association — Training Room 5601 Minnesota Drive Anchorage, Alaska Tuesday, June 20, 2000 — 10:00 a.m. 1. CALL TO ORDER Chairman Eugene Bjornstad called the meeting of the Bradley Lake Hydroelectric Project Management Committee to order at 10:00 a.m. on Tuesday, June 20, 2000, from the Chugach Electric Association’s Training Room, Anchorage, Alaska, to conduct the business of the Committee per the agenda and public notice. 2. ROLL CALL Roll was called by Shauna Dean. The following members were present: Gene Bjornstad Chugach Electric Association Wayne Carmony Matanuska Electric Association Norm Story Homer Electric Association George Kitchens Golden Valley Electric Association Stan Sieczkowski Alaska Energy Authority (teleconference) Hank Nikkels Anchorage Municipal Light & Power Dave Calvert City of Seward 3. PUBLI L. Bob Zharsky, Homer Electric Association Don Stead, Homer Electric Association Bob Day, Anchorage Municipal Light & Power Brian Hickey, Chugach Electric Association Dora Gropp, Chugach Electric Association John Cooley, Chugach Electric Association Mike Keech, Anchorage Municipal Light & Power Larry Hembree, Anchorage Municipal Light & Power Mike Mass, Chugach Electric Association Bob Price, Anchorage Municipal Light & Power Tuckerman Babcock, Matanuska Electric Association Steve Haagenson, Golden Valley Electric Association Ron Saxton, Ater Wynne, LLP Dennis McCrohan, Alaska Energy Authority Shauna Dean, Alaska Energy Authority 4. P ENT There were no public comments. SF AGENDA COMMENTS Chair Bjornstad noted that the group is not technically required to have the Election of Officers until after the first meeting in July. He left the Election up to the Committee to decide. Also, on a the Bradley Lake Instability Study, Brian Hickey will make a presentation, which will be followed by a recommendation from the O&D Committee, which will be handled by Stan Sieczkowski. 6. APP’ MEET -M MOTION: Mr. Story moved to approve the meeting minutes of March 24, 2000. Mr. Kitchens seconded the motion. A voice vote was taken and the minutes were unanimously approved. 7. NEW BUSINESS 7A. ELECT FFI - The Chair opened the floor for nominations of Chairman, Vice Chairman, and Secretary/Treasurer. MOTION: Mr. Story moved to nominate Eugene Bjornstad as Chairman, seconded by Dave Calvert. Mr. Calvert moved to nominate Norm Story as Vice Chairman, seconded by Wayne Carmony, and Secretary/Treasurer remains with the Alaska Energy Authority and will be filled by Stan Sieczkowski. There being no other nominations, the Chair closed nominations. A voice vote was taken, and the motion passed unanimously. \ 7’) 4 Add ober 7B. = Hi (kp NTA x “We have completed our analysis of the problems ae at the plant. BE sovetones three alternatives for you to look at. Those alternatives*were not in a report that went out in your package on the 9". There have been some minor modifications to that report and there are new copies on the table We have completed a preliminary economic analysis of the alternatives.we- have done it a couple different ways, and—texplainte-yeu-hew-we-did-that. We have a recommendation on where we should go from here with the plant. What | want to do first is talk about Bradley and the Kenai Peninsula and a concept called dynamic instability. What you (a cr . ; have at Bradley is a lake and a dam at the top of the mountain, a power tunnel that’s about 3 % . miles longyat the end of the power tunnel you have a device called*needles. yend the needles you have the deflectors and a jet brake and those are devices -withis which control the generator. They exist in the turbinepit, which is an importany Patt of our problem. "Also in the ‘ y turbine pit are the buckets Sie he“devices that drive the the water hits the ind those 2. buckets and spins the shaft in the generator. You've got the turbine shaft, you’ve got a a generator, and then you've got this concept called the deflector vs. power curve. We built a / AC power tunnel that by-passes the riveryt kes water out of the lake and feeds it into a powerhouse down at the bottom of the hill. At Brad! ley Lake this tunnel ie about 3% miles long and 13 feet in diameter and has 1,100 feet of head. A typical Pelton turbine operates off of kinetic energy. What happens at a Pelton turbine is yout 1 inte-the md=a-spherieatvetve-— the water is fired out of the needle with very high kinetic energy and it strikes the bucket right here -//” (2 ._ whisbrspins the generator. There is a device called a deflector which you can barely see but it's a scoop that can go over the front of the needle and deflect the water down into the tailrace so you can deflect the water away from the buckets and slow the machine down. What you have at Bradley is water coming in, goes through the needle, the deflector, if the deflector is active it will drive water down, if it is not active, it will let the ween eebe Bue San and turn around and make electricity. At Bradley, we have six needle: rn ially you Adve two units down there, water comes out of the lake, goes around here, and this is needle one, needle two, needle three, four, five, six, and the same on the second unit. You are looking inte the shaft of the generator so the buckets which are not shown on here would be right there and the water \\ > would strike buckets ‘spin the generator around in that general direction. The deflectors at x * Bradley Lake can go from 0 power out to full power out in 1% seconds. They are the fastest control devices in the railbelt, they are faster than combustion turbines. Bradley can operate in two modes — needle mode or deflector mode. In needle mode, the needles are the primary source of controlling power out to the unit and they are very slow. If you*put the unit into deflector mode, the deflector comes partially into the stream, deflects part of the water and it controls the unit so you can control frequency with the deflectors. When-you sare_in needle mode, negdies—are primary—contret if you have a system disturbance, the deflectors will come into the stream even though you are not in deflector mode — they will come in and slow the unit down. Most of the problems that we are talking about are problems we have experienced in needle mode but the deflectors are active because they come in for over- speed control. Ash ta eto Quartz Creek. The Kenai Power System is,a single 115 kV transmission from Anchorag} : 8) from a stability x standpoint on the Kenai — they-are-at Daves Creek there is a static var compensator which has a power oscillation dampening piece of equipment on it and that is a device that will dampen power swings on the system. There is another one of those at Soldotna and those devices are , tuned to dampen oscillations between Kenai and Anchorage at 1% seconds. They are not tuned to dampen _any oscillations that happen at Bradley Lake. You also have Soldotna 1 located in Soldotna is a 36 MW combustion turbine soon to be moved to Nikiski and then you've got the Bernice Lake plant which has two 25 MW turbines and a third turbine that we don’t run very often. naratietinto-Soldotna bstation and para e © a Generator Control: L 5 stable in power system they require a drooping characteristic, which means that as load increases on a generator, speed must decrease. How much of a decrease is defined by a droop and that number is defined by a percentage. Units with small droops will pick up units with large load. Imagine you have a unit with zero droop — a unit with zero droop for no change in frequency will open gate entirely. Because a zero percent change in frequency gets you full opening of the gate,” ‘Minute change in frequency causes the gate to open fully. That's a concept that is going to be important. Once you have a change in the steady-state system — let's say you're operating at some level and a generator drops off line — the governor responds The first thing you have to understand about generator control is that for generators to operate % fi i Q on cron) they will stop on asfrequency that is not,nominal frequency. If you were running at 60 cycles*you lost a generator, the other generators will pick up the load and they will end at a point that is different than 60 cycles. Then you would have to send a new speed reference to the governor to go to 60 and it would put the frequency back at 60 cycles. After droop control is over, the system is no longer at 60 cycles a second and AGC has to raise the system frequency back to where it was. On the power system, if you lose a generator your system will go under speed, it will slow down because you have a drooping characteristic. On a power system,if you lose load, your system will go over frequency and you have to slow the system down. When the system goes over speed at Bradley Lake, the deflectors cut into the stream even though you were in needle mode. Dynamic Instablty_ disturbed, )sTUuR All mechanical systems have a natural frequency — they will oscillate when they are perturbed _¥/ i Most mechanical systems are designed to have equipment that * will dampen that oscillation or slow that oscillation down. Positive Feedb In a control system, negative feedback is good. It causes the system to drive back to the steady-state. Positive feedback causes the system to drive away from the steady state. We experienced a number of unstable oscillations at Bradley starting in 1991. We had control response problems that caused load shed or blackouts on the Kenai Peninsula. Questions: —_—__—_—_— From a Power System perspective, does the Bradley Project initiate instability, or is Kenai Power System or some other external system the source of the problem? Answers: Se ee Bradley Lake is the source of the instability problems observed on the Kenai Peninsula. It’s a classic case of dynamic instability, and small signal perturbatiens- that initiate the instability are generated by deflector turbulence in the turbine pit. disztalbences Questions: ee en If Bradley is the source of the instability, what control system components contribute to or cause the instability? Answers: es The instability is caused by positive feedback in the deflector control circuits. The positive feedback is a result of a dynamic time delay in the shaft speed transducer combined with time delays in the governor logic and hydraulic servomotor system. ~ Questions: ¥ ) ee ¢ ‘ The time delay/positive feedback problem is aggravated by several factors: = Answers: bind tt ges Zero droop when the deflectors are in stream, in needle mode, when a lack of speed reference input to the deflector circuit in needle mode causing the deflectors to remain in stream indefinitely, non-optimal gains cause exaggerated control response, and non-linear relationship of the deflector verses power curve. ~ Questions: ce Why doesn’t the PT! PSS/E model replicate this problem? Answers: Seas The PTI model did not accurately represent the Woodward 501 governor nor the power tunnel. The EPS model accurately represents the governor and tunnel and the tuning performed following machine testing has allowed us to accurately reproduce the oscillations. Question: Why is Bradley's Control response significantly slower than was predicted in the Bradley commissioning and operating studies? Answer: The answer lies in the needle scheduling logic of the Woodward 501 controller. The algorithm contains needle equalization logic that severely limits response time. In addition, speed error is incapable of scheduling additional needles, thus limiting capacity. Question: What is the maximum response rate of the plant? Answer: For the current governor The response can be as the maximum open close rate via AGC is approximately 8 minutes. fong as 25 minutes for a speed error input. Alternative 1 This involves limiting plant operating output such that the units operate low on their deflector verses power curve (below 30 MW). This implies a two-unit plant limit of 60 MW and a single unit plant limit of 30 MW. Alternative 2 Replace governor electronics to reduce time delays. If possible, dampen turbine pit turbulence, and add external intelligence to optimize gains based on power system conditions. Successful implementation of these concepts can virtually eliminate instability and potentially provide the current 27 MW of non-responsive spin. Alternative 3 In addition to number two, install dual-rate hydraulic servomotors to increase needle speeds. Change needle orifices. Implement a dual mode algorithm, “efficiency” verses “system support.” Efficiency mode would maintain needle equalization and system support mode would open six needles as fast as possible. As far as cost is concerned, early in the project we estimated a worst case cost for replacement of the governor, servomotors;and needle orifices at approximately $5M. We have used this value as a worst case cost to‘perform a net present value analysis on each of the alternatives. REQUEST: The Bradley Lake Operations and Dispatch Subcommittee recommends that the Bradley Lake Project Management Committee fund Proof of Concept designs and studies for alternative three in the amount of $856,200. * e * This number represents a correction in the amount of $60,000 over the total request for the estimated cost of the Droop Coordination Study found in the draft final report dated June 9, 2000. MOTION: Mr. Story moved that the BPMC authorize $856,200 from the construction funds for the purpose of the schedule module with the understanding that we get quarterly update status reports and that during the first quarter you refine the schedule and by doing so the intention would be that we have some control. Seconded by Hank Nikkels. A roll call vote was taken and the motion passed unanimously. MOTION: Mr. Nikkels moved that the PMC supports a progressive effort to improve the operation and safety of the Bradley plant by studying and enhancing controls and other critical elements of the plant, thereby enhancing the value to the participating utilities. Second by Norm Story. A roll call vote was taken and the motion passed unanimously. Mr. Kitchens expressed his appreciation, thanks, and gratitude to the Operation and Dispatch Committee for the work they have done identifying this problem and suggesting its proposed resolution. 7C. Approve the reallocation of the SVC spares funding from the current $145,000 to 0 - Sieczkowski. MOTION: Mr. Sieczkowski moved to approve the reallocation of the SVC spares funding from the current amount of $145,000 to $0.00. Seconded by Norm Story. When we developed this list for the construction funds, part of the approval done by the PMC was for buying SVC spares for the amount of $145,000 and Chugach Electric was going to head up that effort. Over the years, after numerous discussions, the Bradley O&D committee has discussed the possibility and the need to go ahead with purchasing those SVC spares and have come to the conclusion that it is not necessary and that spares for the SVC should be ‘ purchased as needed, orwheneverit-comes-abeut. Based on that recent information, iMeadap te O2> puoi to reallocate the SVC spares funding from $145,000 to $0.00. A voice vote was taken, and the motion passed unanimously. 8. COMMITTEE COMMENTS A. Next Meeting Date Bjornstad Next meeting date is at the call of the Chair. 9: ADJOURNMENT MOTION: Mr. Kitchens moved to adjourn. Seconded by Mr. Nikkels. There being no objection and no further business of the Committee, the meeting was recessed at 11:45 a.m. BY: Eugene Bjornstad, Chairman ATTEST: Alaska Energy Authority, Secretary ALASKA INDUSTRIAL DEVELOPMENT * AND EXPORT AUTHORITY /= ALASKA im =ENERGY AUTHORITY 480 WEST TUDOR ANCHORAGE, ALASKA 99503 907 / 269-3000 FAX 907 / 269-3044 BRADLEY LAKE PROJECT MANAGEMENT COMMITTEE MEETING AGENDA Tuesday, June 20, 2000 — 10:00 a.m. Chugach Electric Association Training Room 5601 Minnesota Drive, Anchorage, Alaska 1 CALLTO ORDER |~ Bjornstad 2. ROLL CALL (for Committee members) ~~ 3 PUBLIC ROLL CALL (for all others present) | ~ 4. PUBLIC COMMENT .~ 5. AGENDA COMMENTS ~~ 6. APPROVAL OF MEETING MINUTES — March 24, 2000 N 1: NEW BUSINESS 4 naAY A A. Election of Officers — ©o ope Bjornstad ee B. Bradley Lake Instability Study eae Presee Tia | OW Hickey ees 8. COMMITTEE COMMENTS A. Next Meeting Date Bjornstad 9: ADJOURNMENT r ALASKA INDUSTRIAL DEVELOPMENT AND EXPORT AUTHORITY /= ALASKA @@i _=ENERGY AUTHORITY 480 WEST TUDOR ANCHORAGE, ALASKA 99503 907 / 269-3000 FAX 907 / 269-3044 ALASKA ENERGY AUTHORITY/ALASKA INDUSTRIAL DEVELOPMENT AND EXPORT AUTHORITY Public Notice Bradley Lake Project Management Committee Notice is hereby given that the Bradley Lake Project Management Committee will hold a regular meeting on Tuesday, June 20, 2000, at 10:00 a.m. This meeting will be held at the following location: Chugach Electric Association’s Training Room, 5601 Minnesota Drive, Anchorage, Alaska. For additional information contact Eugene Bjornstad, Chairman. The State of Alaska (AIDEA), complies with Title II of the Americans with Disabilities Act of 1990. Disabled persons requiring special modifications to participate should contact AIDEA staff at (907) 269-3000 to make special arrangements. /s/ Alaska Energy Authority Project Management Committee Publish: Monday, June 12, 2000 BRADLEY LAKE PROJECT MANAGEMENT COMMITTEE MEETING MINUTES Via Electronic Media @ Chugach Electric Association — Training Room 5601 Minnesota Drive Anchorage, Alaska Tuesday, June 20, 2000 — 10:00 a.m. a CALL TO ORDER Chairman Eugene Bjornstad called the meeting of the Bradley Lake Hydroelectric Project Management Committee to order at 10:00 a.m. on Tuesday, June 20, 2000, from the Chugach Electric Association's Training Room, Anchorage, Alaska, to conduct the business of the Committee per the agenda and public notice. 2. ROLL CALL Roll was called by Shauna Dean. The following members were present: Gene Bjornstad Chugach Electric Association Wayne Carmony Matanuska Electric Association Norm Story Homer Electric Association George Kitchens Golden Valley Electric Association Stan Sieczkowski Alaska Energy Authority (teleconference) Hank Nikkels Anchorage Municipal Light & Power Dave Calvert City of Seward 3: PUBLIC ROLL CALL Bob Zharsky, Homer Electric Association Don Stead, Homer Electric Association Bob Day, Anchorage Municipal Light & Power Brian Hickey, Chugach Electric Association Dora Gropp, Chugach Electric Association John Cooley, Chugach Electric Association Mike Keech, Anchorage Municipal Light & Power Larry Hembree, Anchorage Municipal Light & Power Mike Mass, Chugach Electric Association Bob Price, Anchorage Municipal Light & Power Tuckerman Babcock, Matanuska Electric Association Steve Haagenson, Golden Valley Electric Association Ron Saxton, Ater Wynne, LLP Dennis McCrohan, Alaska Energy Authority Shauna Dean, Alaska Energy Authority 4. Pp NT There were no public comments. 5. AGENDA COMMENTS Chair Bjornstad noted that the group is not technically required to have the Election of Officers until after the first meeting in July. He left the Election up to the Committee to decide. Also, on the Bradley Lake Instability Study, Brian Hickey will make a presentation, which will be followed by a recommendation from the O&D Committee, which will be handled by Stan Sieczkowski. 6. P. F MEETI INUTES — March 24, 2 MOTION: Mr. Story moved to approve the meeting minutes of March 24, 2000. Mr. Kitchens seconded the motion. A voice vote was taken and the minutes were unanimously approved. 7. NEW BUSINESS 7A. ELE FFICERS — Bjorn The Chair opened the floor for nominations of Chairman, Vice Chairman, and Secretary/Treasurer. MOTION: Mr. Story moved to nominate Eugene Bjornstad as Chairman, seconded by Dave Calvert. Mr. Calvert moved to nominate Norm Story as Vice Chairman, seconded by Wayne Carmony, and Secretary/Treasurer remains with the Alaska Energy Authority and will be filled by Stan Sieczkowski. There being no other nominations, the Chair closed nominations. A voice vote was taken, and the motion passed unanimously. 7B. BRADLEY LAKE INSTABILITY STUDY — Hickey (REPORT ATTACHED) We have completed our analysis of the problems at the plant and developed three alternatives for you to look at. Those alternatives were not in a report that went out in your package on the 9". There have been some minor modifications to that report and there are new copies on the table. We have completed a preliminary economic analysis of the alternatives and have done it a couple different ways. We have a recommendation on where we should go from here with the plant. What | want to do first is talk about Bradley and the Kenai Peninsula and a concept called dynamic instability. What you have at Bradley is a lake and a dam at the top of the mountain, a power tunnel that’s about 3 %2 miles long and at the end of the power tunnel you have a device called needles. Beyond the needles you have the deflectors and a jet brake and those are devices which are used to control the generator. They exist in the turbine pit, which is an important part of our problem. Also in the turbine pit are the buckets attached to the runner and shaft and the water hits the buckets and spins the shaft in the generator. You've got the turbine Page 2 of 7 shaft, you’ve got a generator, and then you've got this concept called the deflector vs. power curve. We built a power tunnel that bypasses the river, takes water out of the lake and feeds it into a powerhouse down at the bottom of the hill. At Bradley Lake this tunnel is about 3% miles long and 13 feet in diameter and has 1,100 feet of head. A typical Pelton turbine operates off of kinetic energy. What happens at a Pelton turbine is the water is fired out of the needle with very high kinetic energy and it strikes the bucket right there which spins the generator. There is a device called a deflector which you can barely see but it’s a scoop that can go over the front of the needle and deflect the water down into the tailrace so you can deflect the water away from the buckets and slow the machine down. What you have at Bradley is water coming in, goes through the needle, the deflector, if the deflector is active it will drive water down, if it is not active, it will let the water hit the bucket, spin and turn around and make electricity. At Bradley, we have six needles (on each turbine). Essentially you have two units down there, water comes out of the lake, goes around here, and this is needle one, needle two, needle three, four, five, six, and the same on the second unit. You are looking at the shaft of the generator so the buckets which are not shown on here would be right there and the water would strike buckets and spin the generator around in that general direction. The deflectors at Bradley Lake can go from 0 power out to full power out in 1% seconds. They are the fastest control devices in the railbelt, they are faster than combustion turbines. Bradley can operate in two modes — needle mode or deflector mode. In needle mode, the needles are the primary source of controlling power out to the unit and they are very slow. If you put the unit into deflector mode, the deflector comes partially into the stream, deflects part of the water and it controls the unit so you can control frequency with the deflectors. In needle mode, if you have a system disturbance, the deflectors will come into the stream even though you are not in deflector mode — they will come in and slow the unit down. Most of the problems that we are talking about are problems we have experienced in needle mode but the deflectors are active because they come in for over-speed control. The Kenai Power System is tied by a single 115 kV transmission from Anchorage to Quartz Creek. There are two important components from a stability standpoint on the Kenai — at Daves Creek there is a static var compensator which has a power oscillation dampening piece of equipment on it and that is a device that will dampen power swings on the system. There is another one of those at Soldotna and those devices are tuned to dampen oscillations between Kenai and Anchorage at 1% seconds. They are not tuned to dampen any oscillations that happen at Bradley Lake. You also have Soldotna 1 located in Soldotna. It is a 36 MW combustion turbine soon to be moved to Nikiski and then you've got the Bernice Lake plant which has two 25 MW turbines and a third turbine that we don’t run very often. Generator Control: The first thing you have to understand about generator control is that for generators to operate stable in power system they require a drooping characteristic, which means that as load increases on a generator, speed must decrease. How much of a decrease is defined by a droop and that number is defined by a percentage. Units with small droop percentages will pick up more load faster than units with large droop percentages. Imagine you have a unit with zero droop — a unit with zero droop for no change in frequency will open its gate entirely. Because a zero percent change in frequency gets you full opening of the gate, minute change in frequency causes the gate to open fully. That's a concept that is going to be important. Once you have a change in the steady-state system — let's say you're operating at some level and a generator drops off line — the governor responds on droop and they will stop at a frequency that is not nominal frequency. If you were running at 60 cycles and you lost a generator, the other Page 3 of 7 generators will pick up the load, and they will end at a point that is different than 60 cycles. Then you would have to send a new speed reference to the governor to go to 60 and it would put the frequency back at 60 cycles. After droop control is over, the system is no longer at 60 cycles a second and AGC has to raise the system frequency back to nominal. On the power system, if you lose a generator your system will go under speed, it will slow down because you have a drooping characteristic. On a power system, if you lose load, your system will go over frequency and you have to slow the system down. When the system goes over speed at Bradley Lake, the deflectors cut into the stream even though you were in needle mode. Synanie] at All mechanical systems have a natural frequency — they will oscillate when they are disturbed. Most mechanical systems are designed to have equipment that will dampen that oscillation or slow that oscillation down. Positive F In a control system, negative feedback is good. It causes the system to drive back to the steady-state. Positive feedback causes the system to drive away from the steady state. We experienced a number of unstable oscillations at Bradley starting in 1991. We had control response problems that caused load shed or blackouts on the Kenai Peninsula. Questions: From a Power System perspective, does the Bradley Project initiate instability, or is Kenai Power System or some other external system the source of the problem? Answers: Bradley Lake is the source of the instability problems observed on the Kenai Peninsula. It’s a classic case of dynamic instability, and small signal disturbances that initiate the instability are generated by the deflector and turbulence in the turbine pit. Questions: If Bradley is the source of the instability, what control system components contribute to or cause the instability? Answers: The instability is caused by positive feedback in the deflector control circuits. The positive feedback is a result of a dynamic time delay in the shaft speed transducer combined with time delays in the governor logic and hydraulic servomotor system. Questions: Why doesn't the PTI PSS/E model replicate this problem? Page 4 of 7 Answers: The original PT! model did not accurately represent the Woodward 501 governor nor the power tunnel. The EPS model accurately represents the governor and tunnel, and the tuning performed following machine testing has allowed us to accurately reproduce the oscillations. Question: Why is Bradley's control response significantly slower than was predicted in the Bradley commissioning and operating studies? Answer: The answer lies in the needle scheduling logic of the Woodward 501 controller. The algorithm contains needle equalization logic that severely limits response time. In addition, speed error is incapable of scheduling additional needles, thus limiting capacity. Question: What is the maximum response rate of the plant? Answer: For the current governor, the maximum open close rate via AGC is approximately 8 minutes. The response can be as long as 25 minutes for a step speed error input. Alternative 1 This involves limiting plant operating output such that the units operate low on their deflector verses power curve (below 30 MW). This implies a two-unit plant limit of 60 MW and a single unit plant limit of 30 MW. Alternative 2 Replace governor electronics to reduce time delays. If possible, dampen turbine pit turbulence, and add external intelligence to optimize gains based on power system conditions. Successful implementation of these concepts can virtually eliminate instability and potentially provide the current 27 MW of non-responsive spin. Alternative 3 In addition to number two, install dual-rate hydraulic servomotors to increase needle speeds. Change needle orifices. Implement a dual mode algorithm, “efficiency” verses “system support.” Efficiency mode would maintain needle equalization and system support mode would open six needles as fast as possible. As far as cost is concerned, early in the project we estimated a worst case cost for replacement of the governor, servomotors, and needle orifices at approximately $5M. We have used this value as a worst case cost to perform a net present value analysis on each of the alternatives. REQUEST: Page 5 of 7 The Bradley Lake Operations and Dispatch Subcommittee recommends that the Bradley Lake Project Management Committee fund Proof of Concept designs and studies for alternative three in the amount of $856,200. * e * This number represents a correction in the amount of $60,000 over the total request for the estimated cost of the Droop Coordination Study found in the draft final report dated June 9, 2000. MOTION: Mr. Story moved that the BPMC authorize $856,200 from the construction funds for the purpose of the schedule module with the understanding that we get quarterly update status reports and that during the first quarter you refine the schedule and by doing so the intention would be that we have some control. Seconded by Hank Nikkels. A roll call vote was taken and the motion passed unanimously. MOTION: Mr. Nikkels moved that the PMC supports a progressive effort to improve the operation and safety of the Bradley plant by studying and enhancing controls and other critical elements of the plant, thereby enhancing the value to the participating utilities. Second by Norm Story. A roll call vote was taken and the motion passed unanimously. Mr. Kitchens expressed his appreciation, thanks, and gratitude to the Operation and Dispatch Committee for the work they have done identifying this problem and suggesting its proposed resolution. 7C. MOTION: Mr. Sieczkowski moved to approve the reallocation of the SVC spares funding from the current amount of $145,000 to $0.00. Seconded by Norm Story. When we developed this list for the construction funds, part of the approval done by the PMC was for buying SVC spares for the amount of $145,000 and Chugach Electric was going to head up that effort. Over the years, after numerous discussions, the Bradley O&D committee has discussed the possibility and the need to go ahead with purchasing those SVC spares and have come to the conclusion that it is not necessary and that spares for the SVC should be purchased as needed. Based on that recent information, the O&D recommends to reallocate the SVC spares funding from $145,000 to $0.00. A voice vote was taken, and the motion passed unanimously. 8. COMMITTEE COMMENTS A. Next Meeting Date Bjornstad Next meeting date is at the call of the Chair. Page 6 of 7 9: ADJOURNMENT MOTION: Mr. Kitchens moved to adjourn. Seconded by Mr. Nikkels. There being no objection and no further business of the Committee, the meeting was recessed at 11:45 a.m. BY: = ZX l ( / Eugene Bjoy stad, Chair Alaska Energy ATTEST: jority, Secretary Page 7 of 7 Agenda Item No. CITY OF SEWARD MATANUSKA ELEC ASSOC CHUGACH ELEC ASSOC HOMER ELEC ASSOC GOLDEN VAL ELEC ASSOC MUNI LIGHT & POWER ALASKA ENERGY AUTHORITY A=4+ OVER 51% B = AEA CONCUR With A BRADLEY PMC VOTING TB ‘Pied, List b ety Fondling ~Requasd # A ho (Coll Ze -A O —o 000 / 01% J oT a ws ET] f14 30% rane A Ey ms ATE MESES 1” [7 -.0] [REY 26% (TI =] Ey D= MAJORITY VOTING METHOD A: Requiring four yeas with 51% of utilities, with no AEA vote: 1) Procedures for scheduling, production and dispatch of project power. 2) Establishment of procedures for use of each purchaser's water allocation (AEA assent required for license requirements). 3) Selection among alternative methods that do not involve AEA for funding required project work. VOTING METHOD B: Requiring 4 yeas with 51% of utilities and AEA concurrence: 1) Arranging operation and maintenance of project. 2) Adoption of budget of annual project costs. VOTE(93Q3/BC5272) 3) Establishment of FY estimated annual payment obligation and schedule of each purchaser. 4) Determination of annual project costs after each FY. 5) Evaluation of necessity for and scheduling of required project work. 6) Determination of appropriate amount of insurance. 7) Adoption of additional minimum funding amounts for renewal and contingency reserve fund above that required by bond resolution. 8) Selection among alternate methods that involve AEA for funding required project work, 9) Adoption or amendment of procedural committee rules (except dispute resolution). 10) Adoption of project maintenance schedules. 11) Determination of rules, procedures and accounts necessary to manage project when no bonds outstanding. 12) Evaluation and approval of optional project work and compensation for such work. 13) Application of insurance claims proceeds not governed by bond resolution. 14) Approval of procedures and any individual utility agreements relating to electric power reserves for project. 15) Approval of consultants. VOTING METHOD C: Unanimous vote by all (including AEA) VOTING METHOD D: Majority vote (including AEA) Election of Officers DATE: BRADLEY PMC VOTING Agenda Item No. ves ass YES NO ABS CITY OF SEWARD 01% me | el MATANUSKA ELEC ASSOC =: 14% [_ | [|_| re CHUGACH ELEC ASSOC 30% fe fe Pepe PTS] HOMER ELEC ASSOC 12% [Seb ese] a ee Pgee GOLDEN VAL ELEC ASSOC 17% a | | fae] MUNI LIGHT & POWER 26% [See Pe ES] PETE] ALASKA ENERGY AUTHORITY PSE ES] Papers Sree] A=4+ OVER 51% B= AEA CONCUR With A C = UNANIMOUS D = MAJORITY VOTING METHOD A: Requiring four yeas with 51% of utilities, with no AEA vote: 1) Procedures for scheduling, production and dispatch of project power. 2) Establishment of procedures for use of each purchaser's water allocation (AEA assent required for license requirements). 3) Selection among alternative methods that do not involve AEA for funding required project work. VOTING METHOD B: Requiring 4 yeas with 51% of utilities and AEA concurrence: 1) Arranging operation and maintenance of project. 2) Adoption of budget of annual project costs. VOTE(93Q3/BC5272) 3) Establishment of FY estimated annual payment obligation and schedule of each purchaser. 4) Determination of annual project costs after each FY. 5) Evaluation of necessity for and scheduling of required project work. 6) Determination of appropriate amount of insurance. 7) Adoption of additional minimum funding amounts for renewal and contingency reserve fund above that required by bond resolution. 8) Selection among alternate methods that involve AEA for funding required project work. 9) Adoption or amendment of procedural committee rules (except dispute resolution). 10) Adoption of project maintenance schedules. 11) Determination of rules, procedures and accounts necessary to manage project when no bonds outstanding. 12) Evaluation and approval of optional project work and compensation for such work. 13) Application of insurance claims proceeds not governed by bond resolution. 14) Approval of procedures and any individual utility agreements relating to electric power reserves for project. 15) Approval of consultants. VOTING METHOD C: Unanimous vote by all (including AEA) VOTING METHOD D: Majority vote (including AEA) Election of Officers Agenda Item No. CITY OF SEWARD MATANUSKA ELEC ASSOC CHUGACH ELEC ASSOC HOMER ELEC ASSOC GOLDEN VAL ELEC ASSOC MUNI LIGHT & POWER ALASKA ENERGY AUTHORITY A=4+ OVER 51% B = AEA CONCUR With A DATE: BRADLEY PMC VOTING se | sel) fees ™ wm me vis woos 01% [2ekeet |] Febere) [oT] 14a [ZT T =] a a 30% [Res] Fees Pee pe | 12% (Bah ee] Per] 17% ESE) Gee) Eee] 26% [ed [ete ake aber Beers Paik D = MAJORITY VOTING METHOD A: Requiring four yeas with 51% of utilities, with no AEA vote: 1) Procedures for scheduling, production and dispatch of project power. 2) Establishment of procedures for use of each purchaser's water allocation (AEA assent required for license requirements). 3) Selection among alternative methods that do not involve AEA for funding required project work. VOTING METHOD B: Requiring 4 yeas with 51% of utilities and AEA concurrence: 1) Arranging operation and maintenance of project. 2) Adoption of budget of annual project costs. VOTE(93Q3/BC5272) 3) Establishment of FY estimated annual payment obligation and schedule of each purchaser. 4) Determination of annual project costs after each FY. 5) Evaluation of necessity for and scheduling of required project work. 6) Determination of appropriate amount of insurance. 7) Adoption of additional minimum funding amounts for renewal and contingency reserve fund above that required by bond resolution. 8) Selection among alternate methods that involve AEA for funding required project work. 9) Adoption or amendment of procedural committee rules (except dispute resolution). 10) Adoption of project maintenance schedules. 11) Determination of rules, procedures and accounts necessary to manage project when no bonds outstanding. 12) Evaluation and approval of optional project work and compensation for such work. 13) Application of insurance claims proceeds not governed by bond resolution. 14) Approval of procedures and any individual utility agreements relating to electric power reserves for project. 15) Approval of consultants. VOTING METHOD C: Unanimous vote by all (including AEA) VOTING METHOD D: Majority vote (including AEA) Election of Officers Bradley Lake PROJECT MANAGEMENT COMMITTEE MEETING frre LO , AGW (Date 7 = ie cS EA - Tyaetnd rf Cbn~ (Location) PLEASE SIGN IN : NAME l ( Q, lAt1r-~ Na —— = 2 Wt LER REPRESENTING eee Ue 0&8 RT SS att Quuu V Me Colon 41 DEA 5 (Gene oraslecd CAugecls s a 6 ae Jak = Pa chisiny vA ps 7 George (Sehens | CveA 8 STEW //AAZ F-Irae) GVEA 9 | pen MLW s |MKZ 10 | Tocknnrwn Bolrcot— rE | Wayue Caron’ MA 12 ae c CEL | Bos fuevce |¢Le P 4| gre Cropp Ce s | “ob Lx, MU 5f 16 | Jo vs Cook exq Chuga da 17 — 19 7" 20 21 = 2 ] 92Q2\IT9884 Bradley Instabil ity S tudy Cee Sra: rene Final Report June 20,2000 Prepared by: Brian Hickey Randy Johnson Larry Hembree The project team has completed ou analysis of the problems associated with the Bradley Lake Plant. We have developed three alternatives to mitigate the problems. We’ve completed a preliminary economic analysis of the alternatives And, we’ve have a recommendation on how to proceed ¢ Review of the Bradley Plant: — Lake & Dam — Turbine Pit — Power Tunnel — Buckets — Needles — Turbine Shaft — Deflectors — Generator — Jet Brake — Deflector vs. Power Curve T ypical plan and profile of a High Head hydro power development Hydroulic broke Actual layout of the Bradley Lake Pelton tumbine ¢ The Kenai Power System: — The Daves Creek-University Line — Daves Creek SVC — Soldotna SVC — Soldotna One — Bernice Lake Power Plant — Soldotna- Bradley Line — Dimond Ridge Line ¢ Generator Control: — Speed Error — Speed Reference — Automatic Generation Control (AGC) — Over Speed — Under Speed — Generator Droop 93-28 STANDARD HANDBOOK FOR ELECTRICAL ENGINEERS om suteut L — — position larerncte dwors) feedbock) FIG. 9-25 Schematic diagram of ciectric-hydraulic governor. Dynamic instabéfit ¢ The Narrows Bridge, Tacoma Washingt November 7, 1940 under 40 mph wind load. ¢ Dynamic Electrical Stability: — Steady State Events — Natural frequency — Damped Oscillation — Under Damped Oscillation — Unstable Oscillation ¢ Dynamic Electrical Stability: — Negative Feedback — Positive Feedback — Proportional, Integral, Derivative Gains Why did we begin to sty the ¢ Severe power swings, unstable ose1 at Bradley following minor power system deviations. ¢ Control response problems that led to load shed or blackouts on the Kenai Peninsula. Bradley Lake unit 1 & 2 power swings, 5-24-00 instability Kenai frequency after the University to Daves Creek line trip __W30: BLPP Hz event 048539 vs Bradley Hz event 046260 7/18/96 io Bradley Hz(grn) i | Tiree f i i ii I | 56 | Bradley isolates from Kenai a | i 54.8 Hz ee ia —|— | a + oe Re a BLPP Hwa ~~ | | { 62 1 Z i 200 400 600 800 4000 4200 1400 4600 Bernice Lake Response to University to Daves Creek line trip 7 200 400 600 800 4000 1200 1400 4600 i t | | ee | | A, 7 | ; BLPP 3 MW(red) | | 60 | 35 MW — ‘ill | Leoinereed 4M a | | | clita salen a fe raeping | | | ee “enema ntnaer Sif OY cree | | 0 et ii | | i iit pee | | $5 nt, | | 20 BLPP t2(gmn) ~ | | 10 | ' 50 eeu ig 0 | | 40 Fe eee nea Secale se ad ea eee ee Bernice Lake and Bradley Lake BeSPONSE to University - Daves Creek line trips Zz: W532: Bradiey MW vs BLPP 3 MW 7/18/96 event 0455390452500 = 0 200 400 600 800 4000 4200 4400 4600 30 BLPP 3 MW(gm) Brad ley MW(red) a / 10 wn? 5 o 20 400 600 2~——S—«B OO 1000 1200 1400. ~—«-1600 Questions that we set out to Answer in thesStudly. ¢ Regarding Dynamic Electrical oO: — From a Power System perspective, does the Bradley Project initiate instability, or is Kenai Power System or some other external system the source of the problem ? Questions that we set out to answer in th ¢ Bradley Lake is the source of the instability problems observed on the Kenai Peninsula. ¢ Bradley Lake instability is a classic case of dynamic instability. ¢ The small signal perturbations that initiate the instability are generated by deflector turbulence in the turbine pit, at times aided by the Kenai load variations. Questions that we set out to answer in th — If Bradley is the source of the instability, what control system components contribute to or cause the instability? A. ¢ The instability is caused by positive feedback in the governor deflector controls. ¢ The positive feedback is a result of a dynamic time delay in the shaft speed transducer combined with time delays in the governor logic and hydraulic servomotor system. Transient phase shift in the shaft speed transdue, ieee Questions that we set out to answer in th — The time delay/positive feed back prob ae aggravated by several factors: ¢ Zero droop when the deflectors are in stream, in needle mode. ¢ Lack of a speed reference input to the deflector circuit in needle mode, causes the deflectors to remain in stream indefinitely. ¢ Non-optimal gains cause exaggerated control response ¢ The non-linear relationship of the deflector verses power curve. Questions that we set out to answer in th — Why doesn't the PTI PSS/E model replicate this problem? A. ¢ The PTI model did not accurately represent the Woodward 501 governor nor the power tunnel. ¢ The EPS model accurately represents the governor and tunnel and the tuning performed following machine testing has allowed us to accurately reproduce the oscillations Questions that we set out to answer in thesstudy. ¢ Regarding the Control Response Problem: Q. — Why is Bradley's Control response significantly slower than was predicted in the Bradley commissioning and operating studies? Questions that we set out to answer in th ¢ The answer lies in the needle scheduling logic of the Woodward 501 controller. The algorithm contains needle equalization logic that severely limits response time. ¢ In addition speed error is incapable of scheduling additional needles, thus limiting capacity Questions that we set out to answer in th Q. ¢ What is the maximum response rate of the plant? A. — For the current governor the the maximum open close rate via AGC is approximately eight minutes. — The response can be as long as 25 minutes for a speed error input. What can we do abowutsit: ¢ The “do nothing” option is not an ~ alternative, eventually it will result in severe damage to Bradley Lake and potentially other rotating equipment. What can we do abo ¢ We have developed three alternat — Make no physical modifications and mitigate the instability problem by altering allowable operating limits. — Replace and upgrade electronics and virtually eliminate the instability. — In addition to alternative two, add mechanical equipment to improve control response. e Alternative one — Involves limiting plant operating output such that the units operate low on thier deflector verses power curve (below 30 MW). This implies a two unit plant limit of 60 MW and a single unit plant limit of 30 MW. ¢ Alternative two — Replace Governor electronics to reduce time delays. If possible, dampen turbine pit turbulence, and add external intelligence to optimize gains based on power system conditions. Successful implementation of these concepts can virtually eliminate instability and potentially provide the current 27 MW of non- responsive spin. ¢ Alternative three — In addition to number two, install dual-rate hydraulic servomotors to increase needle speeds. Change Needle Orifices.Implement a dual mode algorithm, “efficiency” verses “system support”. Efficiency mode would maintain needle equalization and System Support mode would open six needles, as fast as possible. What will itee@st? Ch Ree ¢ Early in the project we estimate case cost for replacement of the governor, servomotors and needle orifices at approximately 5 million dollars. We have used this value as a worst case cost to perform a net present value analysis on each of the alternatives. Net Present Va Discounted Cash Flow Analysis Alternative 1 Alternative 2 Alternative 3 Improve Governor Change Operating Improve Governor Controls & Procedures Controls Enhance Response Net Present Value $60.4 $87.3 $95.6 Cost Benefit Cost Benefit Cost Benefit Present Value $0.1 $60.5 $3.0 $90.3 $5.0 $100.6 Year 2000 $0.1 $3.0 $5.0 2001 $4.9 $7.4 $8.2 2002 $5.0 $7.5 $8.4 2003 $5.1 $7.7 $8.6 2028 $8.9 $13.2 $14.7 Subcommittee recommends that the Bradley Lake Project Management Committee fund Proof of Concept designs and studies for alternative three in the amount of $856,200* * This number represents a correction in the amount of $60,000 over the total request for the estimated cost of the Droop Coordination Study found in the draft final report dated June 9, 2000. N VNAT VE Bradley Lake Proof of Concept/Design & Studies 2001 ID | Task Name Duration Start Finish Jul Aug | Sep | Oct | Nov | Dec | Jan | Feb | Mar | Apr | May | Jun Jul Aug | Sep | Oct | Nov 1 | Bradley Proof of Concept/Design & Studies 367 days/ Thu 07/06/00| — Fri 11/30/01 emetic maaan 2 Design & Implementation of Interim Instability Mitigati 8wks| Thu 07/06/00} Wed 08/30/00 Project Team,Dispatchers 3 Enhance/Harden Instrumentation System 6wks| Thu 08/31/00} Wed 10/11/00 | i R. Johnson 4 Governor Control Algorithm Design 25 wks} Thu 10/12/00| Wed 04/04/01 | Project Team 5 Develop Functional Specification 5 wks} Thu 04/05/01} Wed 05/09/01 | | Consultant 6 Identify Source of Turbine Pit Mitigation Devices 12 wis Fri 09/01/00| Thu 11/23/00 | Consultant 7 Design Turnine Pit Mitigation Devices 24 wks Fri 11/24/00} Thu 05/10/01 | c i onsultant - ee : | 8 Railbelt Peak Mode Droop Coordination Design 4wks| Tue 08/01/00} Mon 08/28/00 | ee Machine Rating Subcommittee,Consultant 9 Integrated Exciter Power Sys Stabilizer Design 3wks} Wed 11/01/00} Tue 11/21/00 i Consultant pe 10 Design a Dual Rate Servo Motor System 25 wks| Mon 06/11/01 Fri 11/30/01 | Consultant age ee 11 Design External Intelligence & SCADA Interface 12wks| Tue 05/01/01} Mon 07/23/01 Consultant - SCADA Analyst 12 Design Condense to Generate Logic 8wks| Tue 07/24/01} Mon 09/17/01 | Consultant - SCADA Analyst 13 Design of POD Installation of Soldotna Bradley Lines 12wks| Thu 03/01/01} Wed 05/23/01 ABB Eee 14 Develop Project Implementation & Const Plan 2wks} Thu 11/01/01} Wed 11/14/01 | Consultant Bj 15 Design Fiber Optic Shaft Speed Transducer 8wks| Wed 11/15/00} Tue 01/09/01 Consultant Eee 1 Revew 8 Tesing of Mardiecier Sappied Models eT Gays TOE) OBUIS0 (iki reat Teas ies eee = a ae ae Task ae Milestone * Rolled Up Split sunwwewwewaween &xternal Tasks eae one Proof of Conce Split ‘viwuwwinienwn, summary any Rolled Up Milestone © Project Summary aoe Progress MEE Rolled Up Task 9 [ENN Rolled Up Progress TT Page 1 Bradley Lake Proof of Concept/Design & Studies 2001 ID__|Task Name Duration Start Finish Jun Jul Aug | Sep | Oct | Nov | Dec | Jan | Feb | Mar | Apr | May | Jun Jul Aug | Sep | Oct | Nov 1 |Bradley Proof of Concept/Design & Studies 367 days| Thu 07/06/00 Fri 11/30/01 | en Ty 2 Design & Implementation of Interim Instability Mitigati 8wks| Thu 07/06/00} Wed 08/30/00 Project Team,Dispatchers 3 Enhance/Harden Instrumentation System 6wks| Thu 08/31/00} Wed 10/11/00 R. Johnson 4 Governor Control Algorithm Design 25 wks} Thu 10/12/00! Wed 04/04/01 e Project Team 5 Develop Functional Specification 5wks| Thu 04/05/01} Wed 05/09/01 Consultant 6 Identify Source of Turbine Pit Mitigation Devices 12 wks Fri 09/01/00} Thu 11/23/00 Consultant 7 Design Turnine Pit Mitigation Devices 24 wks Fri 11/24/00} Thu 05/10/01 Consultant 8 Railbelt Peak Mode Droop Coordination Design 4wks| Tue 08/01/00} Mon 08/28/00 ae Machine Rating Subcommittee,Consultant 9 Integrated Exciter Power Sys Stabilizer Design 3wks| Wed 11/01/00} Tue 11/21/00 Consultant Ee 10 Design a Dual Rate Servo Motor System 25 wks| Mon 06/11/01 Fri 11/30/01 Consultant ee eee 1 Design External Intelligence & SCADA Interface 12wks| Tue 05/01/01} Mon 07/23/01 | Consultant - SCADA Analyst 12 Design Condense to Generate Logic 8wks| Tue 07/24/01} Mon 09/17/01 | | Consultant - SCADA Analyst | 13 Design of POD Installation of Soldotna Bradley Lines 12wks| Thu 03/01/01} Wed 05/23/01 | ABB Eee 14 Develop Project Implementation & Const Plan 2wks| Thu 11/01/01} Wed 11/14/01 i Consultant B 15 Design Fiber Optic Shaft Speed Transducer 8wks| Wed 11/15/00} Tue 01/09/01 | Consultant eae 15 | _FevewSTesing fanfare Spas Moses | 357 dve| Thy O7O300/ F770] | RN Task eee Milestone e Rolled Up Split iwwwwannananan &xternal Tasks 2s Project: Bradley Lake Proof of Conce 7" epi easg: Mil Proj Qa EEE Date: Tue 06/20/00 Split Sduuuninnwnann, summary Rolled Up Milestone © roject Summary Progress MEE Rolled Up Task [© Rolled Up Progress Page 1