Talked to Morgan.  The ETS cost estimate is just the Level A cost estimate done by Bryan Reinecke in his Group.  He added that the R.O.W. cost is one of the items has high uncertainty.  Jerry Martin's Group has been not been asked to look at the project.

-----Original Message-----
From: Rosenberg, David E. 
Sent: Saturday, November 17, 2001 12:05 PM
To: Hayslett, Rod
Cc: Tu, Denis
Subject: RE: Toll Model Review Meeting Updates


Suggest asking Mike Smith also.  To me looks like the capex inflation and O&M inflation are backasswards.

-----Original Message----- 
From: Hayslett, Rod 
Sent: Sat 11/17/2001 9:32 AM 
To: Lee, Jebong 
Cc: Tu, Denis; Rosenberg, David E. 
Subject: RE: Toll Model Review Meeting Updates


You should get these 2 experts in pipeline modelling and estimating involved into the discussion.   They have years of experience in inflation and impacts on pipeline economics.

-----Original Message----- 
From: Lee, Jebong 
Sent: Fri 11/16/2001 6:24 PM 
To: Hill, Robert; Gadd, Eric 
Cc: Ratner, Michael; Kissner, Tim; Reinecke, Bryan; Miller, Mary Kay; Hayslett, Rod 
Subject: Toll Model Review Meeting Updates



The model review meeting on Thursday went well, and the consensus was that there are many changes to be made to the current Foothills economic model. Many assumptions in the current model were challenged by the participant companies. Foothills agreed to reflect the suggested changes to the model (Andy is supposed to send out a summary of the discussion) and provided an electronic copy of the model for withdrawn partners' review. Withdrawn partners agreed to give their comments/feedback on the model to Foothills until 11/27/01 and another meeting to reach an agreement on economic model/toll was scheduled for 12/03/01 or 12/04/01. The data from economic/toll model will be used in commercial proposal/presentation to the ANS producers which ANNGTC intends to have by 12/20/01. Tim and I will review Foothills' cash flow and COS model and prepare comments.

Overall, the participants though that the current Foothills model ($0.43/MMBtu toll for Alaska section) has very aggressive assumptions in many ways, and it is very likely that the toll will be increased significantly after all the appropriate revisions are made. The following are the main discussion points of the meeting. Item 1 and 2 are expected to have a very large impact on the toll.

1) Capex and Opex need to have annual/monthly escalation (current model has 0% escalation). For example, Foothills sensitivity analysis shows that 2.5% Capex escalation would increase the toll by $0.20. Numbers such as 1.5% for Capex and 2.5% for Opex were thrown during the meeting, but more reviews need to be done regarding appropriate escalation rates. Do you have any opinion on this, Bryan?

2) Foothills is using Flowthrough method (using cash taxes in COS calculation) to calculate income taxes that go into the Cost of Service and might be underestimating Cost of Service. Tim thought that the Flowthrough method will not fly because of previous cases in which that method was not allowed by IRS and FERC. Foothills said that they can still have same toll under Normalized tax method by changing the depreciation number, but Tim was concerned that it might force the partnership to record negative depreciation. Foothills declined our request to share their Cost of Service model and insisted that we calculate the toll ourselves based on the information in the cash flow model. They agreed to provide a hard copy of their Cost of Service model in the end, but Tim and I were under impression that Foothills was trying to lower the toll as much as they can. Tim was also concerned about levelized toll approach because of FERC might lower allowed return in the later rate cases and more analysis needs to done regradign this. Please correct me if I am wrong here, Tim.

3) OM & A is currently estimated to be 1% of Capex but this might be too low (OM&A in more latest cost estimate by Foothills is approx. 1.5% of total Capex).

4) The Withdrawn Partners' historical contributions ($228 million) will be added to the Rate Base, but this might be dropped if necessary. Success fee of $40 million should be included in the project budget.

5) Working Capital will be added to the project cost. The Line Pack cost (14.5 Bcf for the whole section) was assumed to be borne by ANS producers but some participants highly doubted that that would be the case. The Line Pack gas might have to be bought by the Partnership or the Producers will at least demand certain return on their gas sitting in the pipeline for 25 years.

6) For Return on Capital and AFUDC, 14% ROE and 8% Interest on Debt assuming 30/70 capital structure is likely to be kept. However, the current drawdown schedule assumes construction financing available from the very beginning, which will not be true. Certain assumptions should be made on the financial close date and commercial agreement/preliminary determination date to calculate AFUDC more accurately.

7) Current financing assumption in the model is a 25 year bank loan with balloon amortization and DSCR of 1.6. In reality, the initial bank loans will be of shorter maturity (5-15 years) and certain roll-over financing will be unavoidable. It is also likely that the loans will have multiple tranches with different terms, interest rate, seniority etc. 

8) There is significant currency risk associated with Canadian section of the ANGTS considering that some sources of capital, revenues and part of operating expenses will be in Canadian dollars. A mechanis to mitigate currency risk should be in place and there might be some costs associated with it.

9) Property taxes of 2% in Alaska seems high, but Property taxes during construction period should be added. 

Lastly, there is about $300 million difference ($4,336-$4035) between ETS current cost estimates and the cost estimate in the Foothills model, excluding IDC and AFUDC. Another engineering estimate might have to be dome for larger pipe size (46 or 48 inch) as noted in Bryan's email.  

Thanks, 

JB 

Jebong Lee 
713-853-9722 office 
713-306-8658 cell