Steady-State Analysis of New England’s Interstate Pipeline Delivery Capability

1. Steady-State Analysis of New England’s Interstate Pipeline Delivery Capability Presentation to the NEPOOL Reliability Committee Levitan & Associates, Inc. January 16, 2001

2. Levitan & Associates, Inc. (LAI) Practice Areas • Energy Markets Simulation and Optimization Modeling • Merchant Generation Economics • Pipeline Transportation Management • Fuel Supply Procurement • Power System Engineering/Heat Balance Optimization • ISO Interconnection Policy and Pricing • NUG Contract Administration (Reformation and Buyouts) • Environmental Compliance Strategy • Litigation Support

3. Confidentiality • ISO-NE & LAI have and will continue to comply with the NEPOOL Information Policy - Rev 3, dated August 10, 2000 • Proprietary Information kept Confidential • Steady-State hydraulic model developed from interstate pipeline public domain information • FERC 567 Reports & FERC Flow Diagrams Reflecting Peak Day Design

4. Steady-State Analysis of New England’s Interstate Pipeline Delivery Capability • What the Study Is & Is Not • Individual & Consolidated Models • Steady-State Perspective • No Temporal & Force Majeure

5. Why a Gas Study ? • In 1991, formation of the New England Gas/Electric Discussion Group to examine regional coordination issues between the gas & electric industries • Three objectives of the “Discussion” Group: • to examine the operational reliability of the gas/electric infrastructure • to increase coordination between the industries • to educate industry participants and observers • Analysis includes the modeling of the loss of a major gas & non-gas fired generator, under projected 1995 system conditions • Seven years go by - no similar analysis has been performed since the publication of that EPRI report

6. Purpose of Study • Since 1997, ISO-NE has received applications for interconnection System Impact Studies for almost 40,000 MW of new merchant generation capacity • Virtually all of this newly proposed capacity is advanced combined cycle technology or peaking capacity fueled exclusively by natural gas • There is a need to review the natural gas pipeline infrastructure in New England and its ability to reliably meet the increasing demand of the power generation sector (deliverability study)

7. Scope of Study • Study time frame: 2000 - 2005 • Analysis of existing pipeline infrastructure • Analysis of expected pipeline infrastructure additions • Develop a steady state hydraulic engineering model of the pipeline systems serving the NEPOOL region • Analyze impacts of Reference and High Case natural gas demand consumption • Conduct sensitivity analysis and recommend transient analysis (Phase II of the Study) to be conducted in 2001 • Summarize results and issue report of findings

8. Timeline of Developments • Summer of 1999, ISO-NE and NEPOOL MRPC discuss the initiation of a Gas Study Project • Fall of 1999, ISO-NE looks to retain a consultant who can provide technical assistance in the development of the RFP • Four potential Consultants were identified: • Energy Market Decisions (EMDEC) • Supply Planning Associates (SPA) • National Economic Research Associates (NERA) • Tabors, Caramanis & Associates (TCA) • EMDEC chosen to develop the RFP - Mr. John Meeske

9. Timeline of Developments (cont’d) • During the winter 2000, ISO & EMDEC work to develop a Gas Study RFP • RFP Issued March 9th, 2000 - Sent to 25 vendors • Eight vendors respond to the ISO’s RFP • On-site presentations of proposals by select bidders - narrowed down the “Short-List” to three vendors: • Levitan & Associates (LAI) • Energy Ventures Analysis (EVA) • R.W. Beck • Levitan & Associates was chosen as final vendor • Negotiations with LAI continue into early summer to refine the Scope of Work and agree upon costs

10. Contractual Deliverables • Bi-weekly status reports • Draft report of findings to be issued by December 31, 2000 • Final report of findings to be issued by January 31, 2001 • Steady state database model which runs under the Gregg Engineering WinFlowTM application

11. Summary of Results

12. Steady-State Modeling Results • No pipeline delivery constraints on a peak day in Winter 2000-01 • No summer peak day pipeline deliverability constraints through 2005 • Delivery constraints are apparent in Winter 2003 • Shortfall in gas requirements  1,755 MW out of 7,550 MW assumed • Delivery constraints intensify by Winter 2005 • Shortfall in gas requirements  3,226 MW out of 11,500 MW assumed • Theoretical mitigation potential thru back-up fuel: Winter 2003 - 71 gas-fired units totaling 16,000 MW - 51 of which are dual fueled totaling 9,250 MW Winter 2005 - 75 gas-fired units totaling 18,650 MW - 54 of which are dual fueled totaling 11,500 MW

13. Market Dynamics in New England

14. New England Natural Gas Supply Sources

15. New England Natural Gas Infrastructure • New England’s Major Interstate Pipelines • Iroquois • Portland • Algonquin • Maritimes & Northeast • Tennessee • Existing pipeline delivery capacity = 3.6 Bcf/d. • Daily LNG sendout capability at Everett = 0.450 Bcf/d. • Expansion of 0.60 Bcf/d for 1,550 MW Sithe New Mystic Station, possibly Island End • About 1.4 Bcf/d peak day deliverability behind the citygates • Liquids via truck  100 MMcf/d (0.1 Bcf/d)

16. Western Canadian Gas thru TCPL, Iroquois and PNGTS M&N Eastern Canadian Gas thru M&N PNGTS Tennessee Western Canadian Gas thru Tennessee Iroquois LNG from Algeria and Trinidad Algonquin Gulf Coast Gas thru Algonquin And Tennessee New England’s Interstate Pipelines

17. Primary Interstate Pipelines

18. Interstate Transportation Market Dynamics • 14 pipeline projects placed in-service during 1999-’00 + 2.0 Bcf/d in the Greater Northeast • New Pipelines in New England, M&N and PNGTS, result in + 615 MMcf/d (0.615 Bcf/d), or about 3800 MW • Counterflow capability through Dracut  Tennessee • Pressure and flow benefits improve network reliability • New LNG supplies from Trinidad • Commoditization of the “Supply Chain” • Repackaged Btu services • Synthetics • Increased liquidity • Risk management

19. Assumptions

20. Reference Case High Case Annual Net Energy Growth Rate 1.5% 2.4% Summer Peak Demand Growth Rate 1.7% 2.9% Winter Peak Demand Growth Rate 1.6% 2.5% Capacity Additions 7,551 11,579 Steady-State Demand Assumptions Two Gas Demand Cases developed by ISO-NE & LAI: Reference Case & High Case

21. Electric Assumptions • ISO-NE develops electric side assumptions • PROSYM production simulation model • Analysis performed for 2000 - 2005 • ISO-NE assumptions for: • projected NEPOOL loads, • existing & proposed capacity and capacity attrition • net-interchange with New York, New Brunswick and Hydro-Quebec • ISO-NE delivers hourly gas demands for NEPOOL units for peak day (summer/winter) and 60 day winter average (Dec 15th thru Feb 15th)

22. Electric Assumptions - Reference Case • Reference case load growth forecast thru 2005 • 7,500 MW (winter) of new capacity by 2005 • 200 MW of capacity attrition - 2000 CELT Report • Net Interchange: • firm contracts per 2000 CELT Report - (NY, NB, HQ) • modeling of post-HQ FEC deliveries - (HQ Phase II) • modeling of NEPOOL sales via proposed new interconnections (cross-sound cable)

23. Electric Assumptions - High Case • High case load growth forecast thru 2005 • 11,500 MW (winter) of new capacity by 2005 • 4,000 MW (winter) of capacity attrition • Net Interchange - Higher than Reference case: • firm contracts per 2000 CELT Report - (NY, NB, HQ) • modeling of post-HQ FEC deliveries - (HQ Phase II) • modeling of NEPOOL sales via proposed new interconnections (cross-sound cable & Bridgeport cable)

24. Merchant Entry in New England (High Case)

25. Merchant Entry by Pipeline (2005 High Case)

26. Merchant Entry by Pipeline

27. Forecast of Annual LDC Gas Demand

28. Monthly Gas Send-Out in New England Source: NEGA, 2000 Statistical Guide, March 2000

29. LNG/Propane Storage Withdrawals Pipeline Storage Injections Typical New England LDC Daily Gas Send-Out Source: WEFA, Northeast Natural Gas Markets, Opportunities and Risks, November 1998

30. Load Profiles and Seasonality • Winter • Reliance on Peak Day System Flow diagrams from various certificate applications to serve merchant generators • Summer • Statistical inference from LDCs’ normalized sales

31. Gregg’s WinFlow Steady-State Model • WinFlow is a shell, requiring extensive and elaborate customization • WinFlow calculates the balanced steady-state pressure-flow relationships for pipeline networks

32. Validation of Steady-State Models • Each individual interstate pipeline model matched its Peak Day Flow diagram within industry tolerances •  5# psi •  10 hp • Steady-state models for Algonquin, Tennessee and M&N were reviewed and informally validated with individual pipelines

33. Scheduling Priorities during Constraints • Primary Firm Transportation  LDCs, to a lesser extent, QFs and some merchants • Secondary Firm Transportation (quasi-firm)  Marketers and merchant generators • Interruptible Transportation  Industrials, merchant generators

34. Findings and Observations

35. * 6970 Btu/kWh 2001 2003 2005 Projected Shortfalls in Gas Requirements (MW)*

36. Summary of Peak Day Gas Scenarios – Total Regional Demand vs. System Capacity

37. Steady-State Modeling Results Unserved merchant capacity does not take into account back-up fuel capabilities.

38. Back-up Fuel Issues • Infrastructure • Air Permits • Delivery Logistics • Tankage • Refill • Operational Constraints, e.g. switch-on-the-fly • Sustainability

39. Mitigation Potential

40. Electrical Contingency: Loss of Major Gas-Fired Generating Unit • No significant loss of pressure or flows • Interstate pipelines have the ability to divert and/or re-route gas along the 1100-mile transportation path

41. Electrical Contingency: Loss of 2000 MW Hydro-Quebec Phase II Facility • Winter Peak Day (after winter 00/01) - System cannot transport any additional gas • Summer Peak Day - More than sufficient pipeline capacity to support replacement gas fueled generation

42. Gas Contingency 1 • Increased horsepower requirements at other compressor stations • Fall in delivery pressures to levels that could disrupt plant operations • No observed impact on other pipelines • Available compression capacity at Burrillville on Algonquin derated from 11,400 hp to 5,700 hp

43. Gas Contingency 2 • Downstream compressor stations able to make-up for loss • No unacceptably low delivery pressures for merchant plants observed • No impact on other pipelines • Available compression capacity at Agawam on Tennessee derated from 9,760 hp to 3,253 hp

44. Gas Contingency 3 • Downstream compressors able to compensate for pressure loss • 7 miles of Tennessee’s 36-inch line at NY-MA border removed

45. Recommendations

46. Recommendations • Certify quality of interstate transportation arrangements • Increase understanding of merchant generators’ fuel-switching capabilities • Promote coordination of power and natural gas scheduling protocols • Advocate the streamlining of FERC’s pipeline certification process

47. LAI Project Team • John Pitts Senior Consultant • John Mesko, P.E. Senior Consultant • Lilly Zhu Consultant • Shilpa Shah Assistant Consultant • Richard Levitan President • John Bitler Principal • Edward McGee, P.E. Managing Consultant • Jack Elder, P.E. Manager, Power Systems and Technology

48. Levitan & Associates, Inc.www.levitan.comTel: 617-531-2818Email: rll@levitan.com

49. Questions ?