Illinois Commerce Commission v. Federal Energy Regulatory Commission

In the United States Court of Appeals For the Seventh Circuit ____________________  Nos. 13–1674, –1676, –2052, –2262  ILLINOIS COMMERCE COMMISSION, et al.,  Petitioners,  v. FEDERAL ENERGY REGULATORY COMMISSION,  Respondent.  __________________________    Petitions for Review of an Order of the  Federal Energy Regulatory Commission.  Nos. EL05‐121‐006, ‐008.  ____________________  ARGUED APRIL 22, 2014 — DECIDED JUNE 25, 2014  ____________________  Before CUDAHY, POSNER, and TINDER, Circuit Judges.  POSNER, Circuit Judge. It’s been almost five years since we  remanded  this  case  to  the  Federal  Energy  Regulatory  Com‐ mission.  Illinois  Commerce Commission v.  FERC,  576  F.3d  470  (7th Cir. 2009). The petitioners who persuaded us to remand  the  Commission’s  order,  which  allocated  costs  for  certain  new  high‐voltage  network  transmission  lines  (consisting  of  the transmission lines themselves plus transformers, capaci‐ tors,  and  other  ancillary  equipment—for  simplicity  we’ll  2  Nos. 13–1674, –1676, –2052, –2262  generally refer to the entire facility as a “transmission line”),  are not satisfied with the order that the Commission issued  on remand. For that order reinstated without change the or‐ der that we had vacated.  The petitioners are primarily the midwestern members of  a  Regional  Transmission  Organization  (plus  the  Illinois  Commerce  Commission,  which  essentially  is  appearing  on  behalf of Commonwealth Edison, the largest electrical utility  in Illinois) called PJM Interconnection. A Regional Transmis‐ sion  Organization  is  a  voluntary  association  primarily  of  utilities  that  either  own  electrical  transmission  lines  that  comprise a regional electrical grid or generate electricity that  is transmitted to the customers in the region. The association  operates the grid on behalf of the members.  PJM  has  the  largest  peak  load  (the  amount  of  electrical  power  expected  to  be  provided  for  a  sustained  period  of  above‐average  demand)  of  any  Regional  Transmission  Or‐ ganization—also  the  largest  population  and  the  most  trans‐ mission mileage. Its region stretches east and south from the  Chicago  area  (northeastern  Illinois)  to  western  Michigan,  eastern  Indiana,  Ohio,  Kentucky,  Tennessee,  West  Virginia,  Pennsylvania, New Jersey, Delaware, Maryland, the District  of  Columbia,  North  Carolina,  and  Virginia.  Most  midwest‐ ern  utilities,  however,  belong  not  to  PJM  but  to  an  Inde‐ pendent  System  Operator  (which  is  similar  to  a  Regional  Transmission  Organization,  however)  called  Midcontinent  Independent System Operator, Inc. (MISO). As shown on the  map  (prepared  by  the  ISO/RTO  Council,  www.isorto.org/ about/default,  visited  June  23,  2014,  as  were  the  other  web‐ sites  cited  in  this  opinion),  MISO  operates  in  the  Midwest,  South,  and  some  of  the  Great  Plains  states,  in  contrast  to  Nos. 13–1674, –1676, –2052, –2262  3  PJM,  which  operates  mainly  in  the  mid‐Atlantic  region  but  also, though to a considerably lesser extent, in the Midwest.  The  Federal  Energy  Regulatory  Commission’s  order  is  ad‐ dressed  only  to  PJM,  but  MISO  will  play  a  bit  role  in  our  analysis.    REGIONAL TRANSMISSION ORGANIZATIONS     What we’ll refer to as the western region of PJM consists  of the parts of Michigan, Illinois, and Indiana shown on the  map as being in PJM’s domain, along with all of Ohio. Elec‐ trical  generating  plants  in  the  western  region  usually  are  close to the  customers—Chicago,  for  example,  a major elec‐ tricity market, is ringed by power plants—and so in that re‐ gion relatively low‐voltage transmission lines (typically 345‐ 4  Nos. 13–1674, –1676, –2052, –2262  kilovolts) are adequate for serving most customers, although  the region also has a number of high‐voltage—765‐kV—lines  for  transmitting  electricity  with  greater  efficiency,  mainly  from Indiana to customers in Chicago. The cities in the east‐ ern region use even lower voltage (230‐kV lines) than the cit‐ ies  in  the  western  region,  but  most  of  the  power  plants  are  farther  away  from  the  customers  than  in  PJM’s  western  re‐ gion  and  therefore  500‐kV  lines  are  preferred  even  though  more expensive; the reason is that higher voltage reduces the  amount of electricity that is lost as a function of the distance  over which it is transmitted.  The  question  presented  by  the  petition  for  review  is  the  extent  to  which  the  members  of  PJM  in  its  western  region  (we’ll  call  these  the  “western  utilities”)  can  be  required  to  contribute  to  the  costs  of  newly  built  or  to‐be‐built  500‐kV  lines  (we’ll  call  these  the  “new  transmission  lines”)  even  though  the  lines  are  primarily  in  the  eastern  part  of  PJM.  Originally  at  issue  were  18  such  lines  and  related  projects,  expected to cost $6.6 billion in toto. The number of new lines  has  dwindled  to  12  (11  already  built,  the  other  under  con‐ struction; but 3 more are under study). The current estimate  of  the  total  cost  of  the  projects  that  have  been  or  will  be  completed is $2.7 billion.  PJM’s western utilities are unlikely to obtain a significant  additional  supply  of  electricity  from  the  new  transmission  lines.  The  capacity  of  the  western  utilities  to  generate  elec‐ tricity is already ample—so ample that they transmit a great  deal  of  their  electricity  to  the  eastern  members  of  PJM  to  help them meet the heavy eastern demand for electricity. Be‐ cause the demand for electricity is so much greater in PJM’s  eastern  subregion,  it’s  unlikely  that  much  electricity  will  be  Nos. 13–1674, –1676, –2052, –2262  5  transmitted  from  the  eastern  to  the  western  utilities  via  the  new transmission lines.  Still, the western utilities may benefit from the new high‐ voltage transmission lines in PJM’s eastern region, and to the  extent they do they can be required to contribute to the cost  of  building  the  new  lines.  The  Commission’s  order  that  we  set  aside  five  years  ago  made  no  effort  to  quantify  those  benefits,  however;  instead  it  allocated  the  costs  of  the  new  transmission lines among all the members of PJM in propor‐ tion  to  each  utility’s  electricity  sales,  a  pricing  method  analogous to a uniform sales tax. The Commission acknowl‐ edged  that  this  was  a  crude  method  of  cost  allocation— which  is  to  put  it  mildly,  because  without  quantifying  the  benefits  of  the  eastern  projects  to  the  western  utilities  it  is  impossible  to  determine  what  those  utilities  should  be  charged:  charging  costs  greater  than  the  benefits  would  overcharge  the  utilities,  and  charging  costs  less  than  the  benefits  would  undercharge  them.  The  Commission  de‐ fended its approach by appealing to the difficulty of measur‐ ing the benefits that the western utilities would derive from  the new lines. We considered that a feeble defense. We said  that  “FERC  is  not  authorized  to  approve  a  pricing  scheme  that  requires  a  group  of  utilities  to  pay  for  facilities  from  which  its  members  derive  no  benefits,  or  benefits  that  are  trivial  in  relation  to  the  costs  sought  to  be  shifted  to  its  members.”  576  F.3d  at  476.  We  acknowledged  that  “if  [the  Commission] cannot quantify the benefits to the midwestern  utilities  from  new  500  kV  lines  in  the  East,  …  but  it  has  an  articulable  and  plausible  reason  to  believe  that  the  benefits  are at least roughly commensurate with those utilities’ share  of total electricity sales in PJM’s region, then fine; the Com‐ mission can approve PJM’s proposed pricing scheme on that  6  Nos. 13–1674, –1676, –2052, –2262  basis.” Id. at 477. But the Commission hadn’t met that stan‐ dard either. So we remanded.  Almost  three  years  elapsed  before  the  Commission  is‐ sued  its  order  on  remand.  PJM  Interconnection,  L.L.C.,  138  FERC  61230  (March  30,  2012).  A  year  later  the  Commission  supplemented  the  order  on  rehearing,  PJM  Interconnection,  L.L.C.,  142  FERC  61216  (March  22,  2013),  and  now,  a  year  farther on, the western utilities are back before us, challeng‐ ing the order on remand—which like the order we set aside  prescribes  “a  region‐wide  postage‐stamp  allocation  of  the  costs of new transmission facilities that operate at and above  500  kV.”  PJM  Interconnection,  L.L.C.,  supra,  138  FERC  61230,  ¶ 49. This is FERC‐speak for allocating the costs of the high‐ voltage lines across all the PJM utilities, east or west, in pro‐ portion  to  each  utility’s respective sales. Just as the price of  sending  a  letter  anywhere  within  the  United  States  is  the  same, so the cost that an electrical utility must contribute to a  500‐kV transmission line will, if FERC has its way, be inde‐ pendent of the utility’s location relative to the location of the  transmission line.  The  postal  analogy  is  forced.  Distance  doesn’t  figure  in  the  price  of a  letter, because most  of  the costs  of  postal  ser‐ vice  are  incurred  in  the  postal  facilities  in  which  mail  is  sorted and in local pick‐up and delivery service, rather than  in  the  transportation  of  the  letter  between  distant  locations.  Here we’re  talking about the allocation of the huge costs of  building high‐voltage transmission lines that do not provide  uniform benefits to all the utilities in the region in which the  lines are built.  Much  of  the  Commission’s  order  on  remand  is  devoted  to  hand‐wringing  over  how  difficult  it  is  to  estimate  the  Nos. 13–1674, –1676, –2052, –2262  7  benefits to PJM’s western utilities of the new 500‐kV lines in  the  east  (thus  reprising  its  original  order).  Yet  at  the  same  time the opinion contains detailed dollar estimates of many  of the benefits—but without explaining the basis of the esti‐ mates. Studies are cited from time to time, but the evidence  and analysis on which they’re based are not described. Even‐ tually the Commission threw up its hands and said in its or‐ der on rehearing that “500 kV and above transmission facili‐ ties  provide  a  broad  range  of  benefits,  including  reduced  congestion,  reduced  outages,  reduced  operating  reserve  re‐ quirements, and reduced losses. These benefits radiate from  the  upgraded  facility,  and  thus  are  spread  throughout  the  PJM  region.”  PJM  Interconnection,  L.L.C.,  supra,  142  FERC  61216,  ¶  67  (footnote  omitted).  But  how  far  they  “radiate,”  and how equally, and with what loss of effect as the distance  grows  are  critical  questions  not  answered  in  the  Commis‐ sion’s  order.  The  benefits  may  “spread  throughout”  the  en‐ tire  domain  of  PJM  without  spreading  equally,  or  even  ap‐ proximately equally, among the utilities that comprise PJM.   Of the four types of benefit listed by the Commission in  the passage we just quoted, at least two—reduced electrical  outages and reduced electricity losses—will definitely not be  equally distributed between the utilities in the eastern region  and the utilities in the western region. Outages in the eastern  region  will  be  reduced  because  the  high‐voltage  transmis‐ sion  facilities  will  enable  electricity  to  be  transmitted  with  greater reliability within the region. But outages in the west‐ ern region will be reduced only trivially. The flow of electric‐ ity in PJM’s domain is west to east except there is some flow  the other way from eastern Indiana to the Chicago area. And  the  typical  blackout  or  brownout  occurs  because  of  an  out‐ age in an individual transmission line or transformers, often  8  Nos. 13–1674, –1676, –2052, –2262  because of an overload or weather damage, and the outages  will persist until those lines can be repaired, rather than be‐ ing offset by a new supply of electricity, whether from west  or east.  As  for  reducing  losses  of  electricity  attributable  to  the  distance  over  which  it  is  transmitted,  the  new  high‐voltage  transmission lines will do that in the eastern region because  high voltage is more efficient than low for transmitting elec‐ tricity over long distances. The western  utilities will  benefit  too, because they won’t have to generate as much electricity  to satisfy the eastern demand. And because PJM requires the  western  utilities  to  maintain  reserve  capacity  (just  as  hospi‐ tals  are  required  to  install  generators  to  provide  a  back‐up  supply of electricity should there be an outage) to make up  for  interruptions  in  the  supply  of  electricity  to  the  eastern  utilities, a reduction in those interruptions as a result of the  new  high‐voltage  transmission  facilities  will  enable  the  western utilities to reduce their reserve capacity.  Another  benefit  to  the  western  utilities  will  be  a  reduc‐ tion in congestion in their transmission lines if interruptions  in  transmission  to  the  eastern  utilities  are  reduced  because  transmission lines in the east will be transmitting electricity  at  a  higher  voltage.  Transmission  congestion  occurs  when  customers’  demand  for  electricity  exceeds  transmission  ca‐ pacity, resulting in what is called “curtailment”: the grid op‐ erator does not allow additional supply to enter the grid be‐ cause  it  would  overload  the  lines.  Curtailment  is  costly  to  the  utilities  because  it  means  they’re  producing  electricity  that  cannot  be  sold  to  their  customers  because  it  cannot  be  transmitted to them.  Nos. 13–1674, –1676, –2052, –2262  9  So some of the benefits of the new high‐voltage transmis‐ sion facilities will indeed “radiate” to the western utilities, as  the Commission said, but “some” is not a number and does  not enable even a ballpark estimate of the benefits of the new  transmission lines to the western utilities. Consider two utili‐ ties, one in northern Illinois and one in southern New Jersey,  whose  peak‐load  capacity  is  the  same.  How  likely  is  it  that  they benefit even roughly equally from a new 500‐kV trans‐ mission facility in New Jersey? The New Jersey utility would  obtain  or  deliver  electricity  using  that  facility;  the  Illinois  utility  could  reduce  its  reserve  capacity  slightly  because  it  would  be  less  likely  to  have  to  help  the  New  Jersey  utility  overcome  an  outage,  as  an  outage  would  be  less  likely.  Those  are  not  equivalent  benefits,  though  treated  by  the  Commission  as  equivalent.  The  only  explanation  for  why  it  did that is that having failed to conduct a cost‐benefit analy‐ sis,  it  had  no  basis  for  treating  the  benefits  as  other  than  equivalent.  The western utilities go to the opposite extreme, arguing  that  their  obligation  to  contribute  to  the  cost  of  the  new  fa‐ cilities  should  be  limited  to  the  percentage  of  their  (that  is,  the  western  utilities’)  electricity  that  flows  through  what  is  called a “constrained” transmission facility (one likely to ex‐ perience  an  outage).  This  is  called  the  “distribution  factor”  or  “beneficiary  pays”  approach,  in  contrast  to  the  Commis‐ sion’s  postage‐stamp  approach.  The  western  utilities  ac‐ knowledge that by enlarging transmission capacity the new  facilities in the east will confer a benefit on them by reducing  the  constraint  factor  and  consequent  outage  danger  in  the  western  subregion.  They  assign  a  very  low  dollar  figure  to  this  benefit,  however,  and  the  Commission  has  shown  that  10  Nos. 13–1674, –1676, –2052, –2262  the figure is an underestimate. But it failed to come up with  its own estimate.  One of the attorneys for the utilities remarked at oral ar‐ gument  that  “utility  executives  and  regulators  have  long  struggled  with  how  to  quantify  reliability  benefits.”  If  one  may  judge  from  its  opinions  in  the  present  case,  FERC  has  given up the struggle. But it has done so prematurely, with‐ out demonstrating that even a rough estimate of the benefits  to  be  conferred  by  the  new  eastern  transmission  facilities  is  impossible.  Cost‐benefit  analysis  is  the  standard  method  of  valuing  large  public  or  commercial  projects,  and  is  hardly  alien to the electric power industry. PJM for example in 2011  conducted a cost‐benefit analysis of a $100 million project to  enlarge  a  500‐kV  transmission  line.  It  estimated  costs  and  benefits  over  the  first  15  years  of  the  project’s  life,  dis‐ counted  them  to  present  value  at  an  annual  rate  of  7.7  per‐ cent, determined the ratio of the present value of the benefits  to  the  present  value  of  the  costs at  14.76,  and  approved  the  project.  PJM,  “MEP‐B‐11  Cost/Benefit  Analysis”  3  (Nov.  3,  2011),  www.pjm.com/~/media/committees‐groups/committe es/teac/20111103/20111103‐2011‐market‐efficiency‐analysis‐r esults‐update.ashx.  (On  the  methodology  of  cost‐benefit  analysis  generally,  see,  e.g.,  Cost‐Benefit  Analysis  (Richard  Layard & Stephen Glaister eds. 1994), and for a short intro‐ duction, see Thayer Watkins, “An Introduction to Cost Bene‐ fit  Analysis,”  www.sjsu.edu/faculty/watkins/cba.htm.)  Of  course  it’s  often  difficult  to  obtain  reliable  predictions  of  costs  and  benefits,  as  long  recognized  in  the  extensive  aca‐ demic  literature  on  cost‐benefit  analysis  of  big  public  infra‐ structure  projects  with  long  expected  lives.  See,  e.g.,  Bent  Flyvbjerg,  “Policy  and  Planning  for  Large‐Infrastructure  Projects:  Problems,  Causes,  Cures,”  34  Environment  &  Plan‐ Nos. 13–1674, –1676, –2052, –2262  11  ning B: Planning and Design 578 (2007); Bert van Wee, “Large  Infrastructure Projects: A Review of the Quality of Demand  Forecasts and Cost Estimations,” in id. at 611; Roger Vicker‐ man,  “Cost‐Benefit  Analysis  and  Large‐Scale  Infrastructure  Projects:  State  of  the  Art  and  Challenges,”  in  id.  at  598.  But  the literature does not infer impossibility from difficulty, as  FERC  apparently  does.  Indeed,  cost‐benefit  analysis  has  been used in more difficult cases than this one, for example  where  some  of  the  costs  or  benefits  are  nonmonetary,  see,  e.g.,  John  Rolfe,  “Cost‐Benefit  Analysis—Some  Practical  Ex‐ amples,” www.cqu.edu.au/__data/assets/powerpoint_doc/00 14/23009/Rolfe‐AGSIP‐CBA‐April‐2007.ppt,  or  where  the  costs are impossible to pinpoint but catastrophic risks exist.  See index references to “cost‐benefit analysis” in Richard A.  Posner, Catastrophe: Risk and Response 316 (2004).  We  do  not  suggest  that  postage‐stamp  pricing  is  appro‐ priate only for the postal service. Our concern is with the ab‐ sence  from  the  Commission’s  orders  of  even  an  attempt  at  empirical justification. The Commission assumes—it does not  demonstrate—that  the  benefits  of  the  eastern  500‐kV  lines  are  proportionate  to  the  total  electric‐power  output  of  each  utility, no  matter how remote the utility is from the eastern  projects  that  the  utility  is  to  be  made  to  contribute  to  the  costs of. It is a method guaranteed to overcharge the western  utilities, as they will benefit much less than the eastern utili‐ ties  from  eastern  projects  that  are  designed  to  improve  the  electricity  supply  in  the  east,  though  the  western  utilities  will  derive  an  incidental  consequence  that  the  Commission  hasn’t  tried  to  quantify.  Contrast  our  wind‐power  decision,  Illinois Commerce Commission v. FERC, 721 F.3d 764 (7th Cir.  2013),  which upheld postage‐stamp  pricing of  the  transmis‐ sion lines required to bring western wind‐generated electri‐ 12  Nos. 13–1674, –1676, –2052, –2262  cal power to the MISO utilities. There was evidence that the  lines would not yield highly disparate benefits to the utilities  asked  to  contribute  to  their  costs.  See  id.  at  774–75.  Indeed,  the  Commission  had  determined  that  the  benefits  from  the  new lines would be spread almost evenly across all the utili‐ ties.  Midwest  Independent  Transmission  System  Operator,  Inc.,  133  FERC  61221,  ¶¶ 54–56  (Dec.  16,  2010).  It  made  no  such  determination in the present case; as a practical matter, all it  did was express a hope that things might turn out that way.  As an example of the unreality of that hope, consider the  500‐kV  project  (eventually  abandoned)  called  Branchburg‐ Roseland‐Hudson, which was to be built in New Jersey at an  expected cost of $946 million. PJM refers to “20 thermal and  reactive  reliability  criteria  violations  in  Northern  New  Jer‐ sey,”  and  these  are  the  only  reasons  given  for  the  project.  Under  the  Commission’s  cost  allocation,  only  about  12  per‐ cent  of  the  cost  of  the  project  would  have  been  paid  by  the  two  principal  New  Jersey  utilities,  while  Commonwealth  Edison  would  have  had  to  pay  almost  16  percent  even  though there was no suggestion that it had contributed more  than trivially (1.26 percent was its estimate, though probably  an  underestimate  because  based  on  its  “beneficiary  pays”  analysis) to those thermal and reactive reliability criteria vio‐ lations.  The Commission relied heavily for its postage‐stamp ap‐ proach  on  an  “ISO/RTO  Metrics  Report”  published  in  2011  by the Regional Transmission Organizations and their cous‐ ins  the  Independent  System  Operators.  Only  two  pages  of  the  report,  however,  refer  to  possible  cost  savings  from  PJM’s  plans,  which  include  the  new  500‐kV  projects,  to  im‐ prove its grid. The discussion of those savings is cursory and  Nos. 13–1674, –1676, –2052, –2262  13  conclusional, as where the report says that “by planning for  future  reliability  needs  on  a  region‐wide  rather  than  a  util‐ ity‐by‐utility  or  state‐by‐state  basis,  PJM’s  Regional  Trans‐ mission  Expansion  Planning  (RTEP)  process  helps  focus  on  transmission  upgrades  that  meet  reliability  criteria  and  in‐ crease  economic  efficiency.  Annual  savings:  $390  million.”  Not only are the calculations that yield the $390 million fig‐ ure not disclosed, but there is no indication of how the bene‐ fits  of  the  increased  efficiency  are  likely  to  be  distributed  across  PJM’s region. Some  of  the  savings that the  report at‐ tributes  to  the  new  projects,  such  as  greater  generation  ca‐ pacity, appear to be irrelevant to utilities in the western sub‐ region,  such  as  Commonwealth  Edison,  because  those  utili‐ ties don’t need additional generation capacity; the need is in  the east.  In denying the petition of Dayton Power & Light (one of  the  western  utilities  challenging  the  Commission’s  postage‐ stamp approach) for rehearing of the Commission’s order on  remand, the Commission had repeated the statement in our  opinion that “if [the Commission] cannot quantify the bene‐ fits  to  the  []western  utilities  from  new  500  kV  lines  in  the  east” it can nevertheless reinstate the order that we had va‐ cated if it “has an articulable and plausible reason to believe  that  the  benefits  are  at  least  roughly  commensurate  with”  the  western  utilities’  share  of  electricity  sales  in  the  entire  PJM  region.  PJM  Interconnection,  L.L.C.,  supra,  142  FERC  61216,  ¶ 38.  But  even  the  modest  goal  of  rough  commensu‐ rability  requires  some  effort  by  the  Commission,  as  we  in‐ sisted,  to  quantify  the  benefits.  It  hasn’t  responded  to  that  directive. Instead it says such things as that the western utili‐ ties “will make use of and benefit from” the new eastern 500‐ kV transmission lines. Id. ¶ 37. The Commission doesn’t ex‐ 14  Nos. 13–1674, –1676, –2052, –2262  plain how much use or how much benefit. Instead it points  out unhelpfully that “flows on the transmission facilities that  operate  at  or  above  500  kV  also  can  change  over  time.”  Id.  ¶ 47. Yes, but how likely is such change, when is it likely to  occur,  and  how  great  is  it  likely  to  be?  These  questions  the  Commission ignores.  The Commission refers repeatedly to the fact that 500‐kV  transmission  lines have  an estimated  useful life  of 40 years,  and it emphasizes that much can change over 40 years. That  is  indeed  true—indeed  a  truism—but  again  unhelpful,  as  it  offers  no  insight  into  the  likely  character  or  direction  of  change over that period. A lot of wind blows over the Atlan‐ tic  Ocean,  and  maybe  some  day,  as  the  Commission  notes,  that  wind  will  generate  electricity  for  Chicago,  or  for  that  matter Seattle. There are plans to build a large wind farm in  the  Atlantic  Ocean  off  Cape  Cod.  See  “Cape  Wind  Completing  Geophysical  Surveys;  Aided  by  Four  Massachusetts  Companies,”  May  12,  2014,  www.cape wind.org/node/1751.  But  there  is  nothing  in  the  Commis‐ sion’s  opinions  on  remand  concerning  when  the  wind  farm  (which  is  controversial,  and  has  been  repeatedly  delayed  since  it  was  first  proposed  in  2001,  Katharine  Q.  Seelye,  “Funds  and  New  Timetable  for  Offshore  Wind  Farm  in  Massachusetts,”  New  York  Times,  Feb.  27,  2014,  p.  A16)  will  be built or whether any of its power is likely to be transmit‐ ted to PJM’s western utilities. We don’t see how the prospect  of such a wind farm justifies making Commonwealth Edison  pay more for a transmission facility designed to reduce out‐ ages  in  New  Jersey  than  the  two  primary  utilities  serving  New Jersey are required to pay.  Nos. 13–1674, –1676, –2052, –2262  15  Furthermore, the Commission, underlining what appears  to be an aversion to cost‐benefit analysis, ignores the need to  discount  future  to  present  value  in  order  to  value  a  future  benefit.  Suppose  it  were  certain  (obviously  it  is  not  certain)  that  in  2060  Commonwealth  Edison  will  derive  a  $100  mil‐ lion  benefit  from an  eastern transmission facility completed  in 2020 (and thus reaching the end of its useful life in 2060)  for  which  it  was  charged  $100  million  that  year.  At  a  dis‐ count rate of 5 percent the present value of that future bene‐ fit would be only $14.2 million.  The  Commission  states  that  since  Exelon  now  owns  not  only Commonwealth Edison but also an eastern PJM utility  (Baltimore  Gas  &  Electric),  Exelon’s  “views  of  the  benefits  that  these  subsidiaries  receive  from  the  new  high  voltage  connection  lines  will  change  over  time  as  corporate  struc‐ tures  change,  blurring  distinctions  between  Eastern  and  Western  PJM.”  Id.  ¶ 48.  But  “corporate  structure”  has  noth‐ ing  to  do  with  the  benefits  that  the  two  subsidiaries  will  or  won’t receive from the Commission’s cost‐allocation system.  Exelon will be delighted by the benefits that its eastern sub‐ sidiary  receives  but  distressed  at  the  costs  that  its  western  subsidiary will incur without corresponding benefits.  The  Commission  notes  Dayton  Power  &  Light’s  argu‐ ment  “that  all  of  the  500  kV  and  above  lines  at  issue  are  hundreds of miles away from [Dayton Power & Light’s] sys‐ tem,  and  that  it  would  be  a  near  impossibility  for  lines  lo‐ cated so  far  away  to  provide any meaningful role in reduc‐ ing  the  number  of  momentary  [outages]  or  outages  of  less  than  an  hour  experienced  on  the  Dayton  system.”  Id.  ¶ 58.  (For remember that when electricity is transmitted over long  distances, some of it is lost.) Dayton Power & Light adds that  16  Nos. 13–1674, –1676, –2052, –2262  “neither  it,  ComEd,  nor  AEP’s  [American  Electric  Power’s]  Ohio  subsidiaries  own  any  500  kV  facilities,  yet  these  com‐ panies  do  not  experience  abnormally  high  outage  rates  on  their  transmission  systems.”  Id.  To  this  the  Commission’s  only  reply  is  that  “Dayton  admits  that  the  Western  PJM  zones  received  some  benefit  from  their  integration  into  PJM.”  Id.  ¶ 79.  But  will  any  of  the  benefit  from  the  new  transmission  facilities  be  in  the  western  subregion?  And  if  so,  how  much?  We’d  settle  for  a  rough  estimate.  The  Com‐ mission made no estimate.  In  similar  vein  the  Commission,  while  acknowledging  that  “western  regions  of  PJM  generally  have  sufficient  gen‐ eration,”  quotes  ComEd  as  saying  that  it  “sought  member‐ ship in PJM first of all because of the reliability benefits that  membership would bring … and the most likely source from  which  ComEd  could  import  energy  to  prevent  loss  of  load  during system emergencies is PJM.” Id. ¶ 76. True. But from  where in PJM?  By now it should be apparent that the basic fallacy of the  Commission’s  analysis  is  to  assume  that  the  500‐kV  lines  that  have  been  or  will  be  built  in  PJM’s  eastern  region  are  basically for the benefit of the entire regional grid. Not true;  their  purpose  is  to  address  specific  reliability  violations  in  the  eastern  part  of  PJM.  No  electric‐power  company  would  spend billions of dollars just to improve reliability in the ab‐ sence  of reliability violations that  required fixing.  There  are  bound  to  be  benefits  to  the  entire  grid  and  therefore  to  the  utilities  connected  to  it,  but  they  are  incidental,  just  as  re‐ pairing  a  major  pothole  in  a  city  would  incidentally  benefit  traffic  in  the  city’s  suburbs,  because  some  suburbanites  commute to the city. So they should pay a share of the cost  Nos. 13–1674, –1676, –2052, –2262  17  of repair, but a share proportionate to their use of the street  with  the  pothole  rather  than  proportionate  to  their  popula‐ tion.  The  incidental‐benefits  tail  mustn’t  be  allowed  to  wag  the primary‐benefits dog.  The order on rehearing was approved by a 3 to 2 vote of  the  FERC  commissioners.  Commissioner  Clark’s  dissent  is  particularly  pointed.  He  denies  that  “there  is  sufficient  evi‐ dence or reasoning in the record to find that benefits for util‐ ities  in  the  Midwest  are  even  roughly  commensurate  to  the  costs  incurred under  the postage stamp methodology. Inas‐ much as this is the case, I believe the Commission’s decision  has  largely  ignored  the  [Seventh  Circuit’s]  clear  directive.”  He  notes  that  the  new  “transmission  facilities  were  ap‐ proved to resolve specific anticipated reliability violations in  the  East,  not  to  increase  the  general  system‐wide  benefits  discussed in the Order on Remand or the Order on Rehear‐ ing.”  He  points  out  that  the  Commission  confuses  benefits  from  belonging  to  PJM,  which  accrue  to  all  the  members (a  member who doesn’t benefit quits—this happens from time  to  time),  with  benefits  from  specific  projects,  noting  suc‐ cinctly that “avoiding overloads in northern New Jersey re‐ duces outages first and foremost for those living in New Jer‐ sey.”  We conclude, with regret given the age of this case, that  the Commission failed to comply with our order remanding  the case to it. It must try again. If it continues to argue that a  cost‐benefit  analysis  of  the  new  transmission  facilities  is  in‐ feasible, it must explain why that is so and what the alterna‐ tives  are.  It  has  presented  no  evidence  that  postage‐stamp  pricing is an acceptable, or the only possible, alternative.  18  Nos. 13–1674, –1676, –2052, –2262  We acknowledge that the benefits of the new facilities to  the western utilities may prove unquantifiable because they  depend  on  the  likelihood  and  magnitude  of  outages  and  other contingencies, and that likelihood and that magnitude  may for all we know baffle the best analysts. If the Commis‐ sion  after  careful  consideration  concludes  that  the  benefits  can’t  be  quantified  even  roughly,  it  can  do  something  like  use the western utilities’ estimate of the benefits as a starting  point,  adjust  the  estimate  to  account  for  the  uncertainty  in  benefit  allocation,  and  pronounce  the  resulting  estimate  of  benefits  adequate  for  regulatory  purposes.  If  best  is  unat‐ tainable  second  best  will  have  to  do,  lest  this  case  drag  on  forever.  To summarize, the lines at issue in this case are part of a  regional grid that includes the western utilities. But the lines  at  issue  are  all  located  in  PJM’s  eastern  region,  primarily  benefit  that  region,  and  should  not  be  allowed  to  shift  a  grossly disproportionate share of their costs to western utili‐ ties  on  which  the  eastern  projects  will  confer  only  future,  speculative, and limited benefits.  The petitions for review are granted and the matter once  again remanded to the Commission for new proceedings.  Nos. 13–1674, –1676, –2052, –2262 19 CUDAHY, Circuit Judge, dissenting. The issues presented here are practically identical with those that we dealt with in Illinois Commerce Comm’n v. FERC, 576 F.3d 470 (7th Cir. 2009)(“Illinois Commission I”). I filed a dissent in that case and I emphatically reiterate its contents here. The majority has expressed a need for more precise numbers about benefits, burdens and a variety of other as- pects. Now it has enhanced that need by suggesting the use of cost-benefit analysis (a method, some think, of dressing up dubious numbers to reach more impressive solutions). I will say preliminarily that I think the majority is under the impression that somehow there is a mathematical solution to this problem, and I think that this is a complete illusion. De- spite the frequency with which cost-benefit analysis is used, it does not resolve the difficulty of accurately or meaningful- ly measuring the costs and benefits involved with these grid strengthening projects. Cost allocation, particularly at these extraordinarily high voltages, is far from a precise science, and there are no mathematical solutions to determining ben- efits region by region or subregion by subregion. See PJM Interconnection, L.L.C., 142 FERC ¶ 61,216 (2013) (“Remand Rehearing Order”)(noting the difficulty of precisely quanti- fying future benefits); see also Illinois Commerce Comm’n v. FERC, 721 F.3d 764, 774 (7th Cir. 2013) (“Illinois Commission II”)(same). Both parties acknowledged this much at argu- ment. Cost allocation is a judgmental matter and should be treated as such. E.g., Alabama Elec. Co-op., Inc. v. FERC, 684 F.2d 20, 27 (D.C. Cir. 1982) (explaining the cost causation principle in a different context). Cost allocation produces approximate results and requires selection of the most ap- propriate methodology among many, none of which are nec- essarily “right.” This is one reason courts should generally 20 Nos. 13–1674, –1676, –2052, –2262 be deferential to FERC’s technical analysis; and, I think somewhat heretically, because the majority’s notions of cost- causation and related technical concepts were not developed in a context of extra-high voltage projects forming a back- bone framework, judicial precedents involving radically dis- tinguishable arrangements, especially those involving lower voltages, are dubious guides to developing an appropriate methodology here. In addition, the majority indulges in descriptions of many elements of the PJM grid and their functions without reference to any engineering evidence in support. For exam- ple, the majority claims that “the cities in the eastern region (of PJM) use even lower voltage (230 kv lines) than the cities in the western region, but most of the power plants are far- ther away from the customers than in PJM’s western region and therefore 500 kv lines are preferred even though more expensive; the reason is that higher voltage reduces [line loss].” Such a statement is at best a vast oversimplification, and the comment that “it’s unlikely that much electricity will be transmitted from the eastern to the western utilities via the new transmission lines” is based on ignoring the poten- tial for future developments of generation and transmission. In fact, the entire thrust of the majority is toward precise cost causation, even in the present case, where that is inde- terminate or at least obscure. The effect of the majority opin- ion is to emphasize functional relationships of the fragments of PJM rather than its value as a unique whole. I do not agree with the majority (or the Commission) that postage stamp cost distribution is “crude.” The reason ascribed by the majority for this deficiency assumes that some other methodology, like DFAX, can trace the benefits of additions Nos. 13–1674, –1676, –2052, –2262 21 with precision—an ability convincingly rejected by the Commission. In fact, the postage stamp methodology is the only one that can be mathematically verified. Thus, if one knows the total cost of the improvements and the total amount of the electrical output, one divided by the other provides an unarguable dividend representing the uniform burden of the various segments. Other methodologies pro- vide approximations, but no more. The majority cites Illinois Commerce Comm’n v FERC, 721 F.3d 764,774 (7th Cir. 2013), the “wind power decision,” as evidence of tolerance for postage stamp allocation but fails to indicate why that decision is not more broadly precedential for this one. In an elaborate effort to distinguish the very similar wind power decision, the ma- jority underestimates the role of a ultra high-voltage back- bone in equalizing benefits for all grid members. Why should not uniformity of benefit as provided by the postage stamp approach be the starting point in both cases? In its critical analysis of an abandoned project in New Jersey the majority cites the alleged single reason for build- ing the project (rather than benefits derived from it). The ma- jority then, by recourse to a Distribution Factor Analysis (DFAX) approach, claims that the New Jersey utilities have been virtually unscathed while Commonwealth Edison has been grossly overcharged. Since this example does not even purport to measure respective benefits (but focuses on mo- tives for construction), I am afraid that it compares apples to (abandoned) oranges. The majority apparently seeks to compare an a priori reason for building the line with benefits (a posteriori) derived from it. In the next paragraph the majority repeats that our earli- er opinion asked for an “articulable and plausible reason” to 22 Nos. 13–1674, –1676, –2052, –2262 believe that certain benefits exist, but rejects the Commis- sion’s efforts to provide one for alleged lack of obviously ob- scure detail. This goes far beyond the proper scope of judi- cial review. The majority derogates the prospect of harness- ing ocean winds, minimizing the well-known efforts to es- tablish a wind farm in the Cape Cod area on the grounds that that transmission project (like many, many others) is controversial. More importantly, the majority seems to de- value future impacts of projects lasting for a half century by improperly discounting future benefits. I could go on reciting in the case of Dayton Power and Light the drumbeat for “precision,” which is simply beyond human capability. I have the impression that the majority is charging the Commission with lack of commitment in pur- suing a “two plus two equals four” solution, but the Com- mission is dealing with incommensurable forces and condi- tions as skillfully and honestly as it can. It has my sympathy as well as my respect. The majority casually concedes the central point that Commonwealth Edison joined PJM for the dominant reason of improving its reliability, but in its unremitting pursuit of fragmentation it insists on identifying exactly the source of the reliability instead of recognizing PJM as an extraordi- narily sophisticated centrally dispatched unit acting as a whole. Nowhere does Commonwealth Edison, in its pursuit of reliability, request a strengthening of some part of the grid, but apparently relies on the reliability that the entire grid provides. It should be no surprise that the Commission split on how to respond to the demands of the majority for more and more precision—specious or otherwise—and in the end the majority concedes that its demand for numbers may be Nos. 13–1674, –1676, –2052, –2262 23 unobtainable and we may have to accept whatever the Commission can produce–whether second best, third best, or whatever. The majority even approves rejection of DFAX, but this was the very basis on which the protesters brought this lawsuit. The only inescapable requirement of the majori- ty seems to be finality and an end to litigation; in that respect I certainly agree with the majority. At one point the majority complains because the Com- mission fails to specify the degree of “radiation” from an upgraded facility and then recites with apparent authority a difference in benefits and lack of uniformity between the eastern and western utilities. Much of this is an effort to supply various details of electrical phenomena, much more the business of the Commission than of this court. The ma- jority seems willing to pursue these details, as speculative and unsupported conclusions, and faults the Commission for not attaching precise magnitudes to its own bottom lines. This is not judicial review; it is manufacturing its own “evi- dence” as a substitute for the Commission’s but still seeking greater exactitude. In any event, since the majority seems to feel free to second guess the Commission, I will indulge in the same freedom (hopefully without too much violence to the Chenery principle) in proposing my own rationale for upholding the FERC proposal. First of all, I think it makes a great deal of sense to start with the presumption that the costs of these extraordinarily high voltage lines ought to be allocated on a shared cost or postage stamp or Asocialized@ basis. 1 These extraordinarily Indeed, the majority in Illinois Commission I acknowledged that a pre- 1 sumption of grid-wide benefits is appropriate for new projects because the grid is integrated and the benefits are inherently spread over the en- 24 Nos. 13–1674, –1676, –2052, –2262 high voltages are not commonly found in electrical transmis- sion systems generally and, if they are found, they constitute what seems to be the backbone of an electrical grid. See PJM Interconnection L.L.C., 138 FERC ¶ 61,230 (Mar. 30, 2012) (“Remand Order”). By that, I mean that they are capable of transmitting large quantities of bulk power long distances to make the entire grid more reliable and more efficient. Id. By their very nature their impacts are broader both geograph- ically and temporally; that is, it=s much more difficult to pin- point their exact benefits to some other part of the system or to confine them to what is going on today rather than tomor- row. See Remand Rehearing Order ¶ 67. In this proceeding, the Commission considered the application of cost allocation by Distribution Factor Analysis (DFAX) but rejected this ap- proach for, I think, adequate reasons 2Bessentially that precise tire grid. 576 F.3d at 477 (“[FERC] can presume that new transmission lines benefit the entire network by reducing the likelihood or severity of outages.”); see also Entergy Servs., Inc. v. FERC, 319 F.3d 536, 543 (D.C. Cir. 2003)(noting that upgrades intended to preserve reliability are pre- sumed to benefit the system as a whole). Given the backbone nature of these extra high-voltage projects, I think this presumption is even more compelling. Under the DFAX methodology, PJM selects the single most severe reli- 2 ability violation for each project, models it on a 5-year period, and does not revisit the allocation even if changes occur to the system before or during construction. See Remand Order ¶ 41, 44. This method also bases allocations solely on who causes the need for the new project, and does not consider who will use the new line once it is built. See id. In other words, the DFAX method is based on limited and temporally inflexible information. Conversely, FERC found that the regional allocation meth- od was adjusted every year to account for changes to the system, as well as use a 15-year projection which allows for greater planning for future developments affecting the grid. See id. ¶¶ 97, 117. Nos. 13–1674, –1676, –2052, –2262 25 benefits and temporally limited impacts were impossible to determine with this approach. See Remand Order ¶ 37. On the other hand, the petitioners here base their entire case up- on an application of this methodology. This is the essence of the case; I think the so-called DFAX methodology is appro- priate for relatively lower voltage transmission, but it is un- suitable for the extraordinarily high voltage backbone for the reasons I have mentioned above. Id. If we start with a presumption that the cost of extraordi- narily high voltage transmission lines should be allocated on a shared cost or postage stamp basis, I see no reason to de- part from that presumption here. The reasons on which the protesters rely are based on an application of the DFAX methodology, which the Commission has found to be inap- propriate for reasons that no court should intervene to reject. See id. ¶¶ 36–47. In fact, I think that the rejection here illus- trates the dangers of substituting the court=s findings in these technical matters for the Commission=s. In general, my view of starting with a presumption (rebuttable, of course) of shared costing for extraordinarily high voltage lines corre- sponds to the reality of the situation reflecting why the line is there and what its basic function is, namely to make the entire grid more functional. 3 See id. ¶ 21. The protests here Reliability is not a middling concern—power outages and the more se- 3 rious “cascading” outages are not uncommon. In 2003 a cascading out- age in Ohio spread across several states, left over 50 million people and resulted in economic losses in the billions. E.g., Mike Edmonds, 10 Years Later, Power Outages Still Cost U.S. Billions Each Year, GridTalk (Aug. 14, 2013), available at http://www.sandc.com/blogs/index.php/2013/08/10- years-later- power-outages- still-cost-u-s-billions-each-year. Estimates put the annual cost of outages upwards of $80 billion. See Kristina Hamachi LaCommare & Joseph H. Eto, Understanding the Cost of Power 26 Nos. 13–1674, –1676, –2052, –2262 are based on the approach of measuring the impact of the backbone network and attempting to target its broad effect to some subregion of the grid. As a matter of equity, many tears have been shed here over the plight of Dayton Power & Light, and particularly, Commonwealth Edison (which, as the power supplier of the forum is singled out), but these utilities joined the system only a few years ago, even though they are many hundreds of miles away from the original PJM (which has been in business for many, many years and was unusually sophisti- cated as a centrally dispatched grid). Interestingly, neither Commonwealth Edison nor its parent Exelon is a party to this proceeding, which may reflect a less intense degree of objection to the outcome. Commonwealth Edison was well aware of the reliance on ultra high voltage transmission as a basic element in improving reliability in PJM, and Com- monwealth Edison is significantly quoted by the majority as having been motivated to join PJM by its need for reliability benefits. Perhaps, Commonwealth Edison was surprised that the cost of any additions would follow a postage stamp basis rather than a DFAX basis, but I doubt that that possibility escaped them completely and I see nothing inequitable about requiring them to participate in the costs of these ad- ditions, which are basically for the benefit of the entire grid. Interruptions to U.S. Electricity Consumers, Lawrence Berkeley National Laboratory, (Sept. 2004), available at http://certs.lbl.gov/pdf/ 55718.pdf. Experts have estimated that the reliability savings from strengthening the transmission backbone, and thus the entire grid, could be as much as $49 billion, annually. See Massoud Amin, U.S. Electrical Grid Gets Less Reliable, IEEE Spectrum (Dec. 30, 2010), available at http://spe- ctrum.ieee.org/energy/policy/us-electrical-grid-gets-less-reliable. Nos. 13–1674, –1676, –2052, –2262 27 This is the essence of my difference with the majority, which says the Commission’s “basic fallacy” is to assume that the 500 kv lines are for the benefit of the entire grid. I do not think this is a fallacy, and even Commonwealth Edison seems to recognize that reality. Commonwealth Edison came late to the party, and I think it is not unfair that they partici- pate on a pro rata basis in these developments. The majority also seems to be totally convinced of the po- sition that, since the electrical flows at the moment are pri- marily from east to west, almost the entire burden should be placed on the eastern utilities; and in that regard my col- leagues were very skeptical of the development of off-shore wind farms in the Atlantic Ocean or other such future devel- opments in electrical generation and transmission, as a pos- sible basis for reversing the flow of power in the future. I can only guess the specific prospects of offshore wind farms or other developments, or how these specific developments might affect the situation, but I certainly don=t think it is un- likely that there will be significant changes in electrical flows over these new facilities during the next 40 or 50 years when the facilities will be in operation. Confining the Commission to the DFAX methodology, which substantially restricts con- siderations of grid development over time, the protesters’ approach ignores the function and outlook of a high voltage backbone. See Remand Order ¶¶ 39, 41, 43, 111. I suppose that the next version of things that we may get from back from FERC will be some sort of hybrid system, which would bow in the direction of what I think should be presumptive (namely, a postage stamp methodology for these extremely high voltage facilities), while maintaining some aspects of an approach that superficially conforms to 28 Nos. 13–1674, –1676, –2052, –2262 various radically distinguishable judicial precedents. In my opinion this will be unfortunate, since I firmly believe we should allow the FERC to be creative in addressing these unprecedented problems. For these reasons I respectfully dissent.