Message-ID: <32071640.1075862267337.JavaMail.evans@thyme> Date: Tue, 20 Nov 2001 05:30:06 -0800 (PST) From: denis.tu@enron.com To: e..rosenberg@enron.com, rod.hayslett@enron.com Subject: RE: Toll Model Review Meeting Updates Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: quoted-printable X-From: Tu, Denis X-To: Rosenberg, David E. , Hayslett, Rod X-cc: X-bcc: X-Folder: \RHAYSLE (Non-Privileged)\Hayslett, Rod\Deleted Items X-Origin: Hayslett-R X-FileName: RHAYSLE (Non-Privileged).pst 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.=20 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-----=20 From: Hayslett, Rod=20 Sent: Sat 11/17/2001 9:32 AM=20 To: Lee, Jebong=20 Cc: Tu, Denis; Rosenberg, David E.=20 Subject: RE: Toll Model Review Meeting Updates You should get these 2 experts in pipeline modelling and estimating involve= d into the discussion. They have years of experience in inflation and imp= acts on pipeline economics. -----Original Message-----=20 From: Lee, Jebong=20 Sent: Fri 11/16/2001 6:24 PM=20 To: Hill, Robert; Gadd, Eric=20 Cc: Ratner, Michael; Kissner, Tim; Reinecke, Bryan; Miller, Mary Kay; Haysl= ett, Rod=20 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 co= mpanies. Foothills agreed to reflect the suggested changes to the model (An= dy is supposed to send out a summary of the discussion) and provided an ele= ctronic copy of the model for withdrawn partners' review. Withdrawn partner= s agreed to give their comments/feedback on the model to Foothills until 11= /27/01 and another meeting to reach an agreement on economic model/toll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 ANN= GTC 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/MM= Btu toll for Alaska section) has very aggressive assumptions in many ways, = and it is very likely that the toll will be increased significantly after a= ll the appropriate revisions are made. The following are the main discussio= n points of the meeting. Item 1 and 2 are expected to have a very large imp= act 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% f= or 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 op= inion on this, Bryan? 2) Foothills is using Flowthrough method (using cash taxes in COS calculati= on) to calculate income taxes that go into the Cost of Service and might be= underestimating Cost of Service. Tim thought that the Flowthrough method w= ill 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 No= rmalized tax method by changing the depreciation number, but Tim was concer= ned that it might force the partnership to record negative depreciation. Fo= othills declined our request to share their Cost of Service model and insis= ted that we calculate the toll ourselves based on the information in the ca= sh 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 tr= ying to lower the toll as much as they can. Tim was also concerned about le= velized toll approach because of FERC might lower allowed return in the lat= er rate cases and more analysis needs to done regradign this. Please correc= t me if I am wrong here, Tim. 3) OM & A is currently estimated to be 1% of Capex but this might be too lo= w (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 (1= 4.5 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 Pac= k gas might have to be bought by the Partnership or the Producers will at l= east demand certain return on their gas sitting in the pipeline for 25 year= s. 6) For Return on Capital and AFUDC, 14% ROE and 8% Interest on Debt assumin= g 30/70 capital structure is likely to be kept. However, the current drawdo= wn schedule assumes construction financing available from the very beginnin= g, which will not be true. Certain assumptions should be made on the financ= ial close date and commercial agreement/preliminary determination date to c= alculate AFUDC more accurately. 7) Current financing assumption in the model is a 25 year bank loan with ba= lloon 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 b= e unavoidable. It is also likely that the loans will have multiple tranches= with different terms, interest rate, seniority etc.=20 8) There is significant currency risk associated with Canadian section of t= he ANGTS considering that some sources of capital, revenues and part of ope= rating expenses will be in Canadian dollars. A mechanis to mitigate currenc= y 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 con= struction period should be added.=20 Lastly, there is about $300 million difference ($4,336-$4035) between ETS c= urrent cost estimates and the cost estimate in the Foothills model, excludi= ng IDC and AFUDC. Another engineering estimate might have to be dome for la= rger pipe size (46 or 48 inch) as noted in Bryan's email. =20 Thanks,=20 JB=20 Jebong Lee=20 713-853-9722 office=20 713-306-8658 cell