Gas Appliance Manufacturers Association, Inc. v. Department of Energy

See, gays and alleged. lesbians is e.g., Doe v. For those reasons because I believe Gates, 981 F.2d 1316 injustice serious has been done to Krc in panel’s uphold decision to summary judg- denying him a trial his claim of agency government ment for the sig- discrimination, this record I grant suggestion nals that this court views such claims inhos- for rehearing en banc. indeed, pitably; it is hard to see how a plaintiff

homosexual will ever obtain a trial

on the issue of whether he was fired or

demoted because of sexual orientation if the

mode of applied persists. review here panel affirms an of summary award

judgment the, on an issue on govern- ment bore the proof ie., %dtimate burden of — whether the USIA would have terminated GAS APPLIANCE MANUFACTURERS regardless Krc of his sexual-orientation —de- ASSOCIATION, INC., al., et spite the lack of contemporaneous USIA Appellants, identifying documents “poor judgment” Kre’s (the primary “neutral” consideration relied

upon) as a sufficient cause for his termi- nation; despite key agency.memorandum DEPARTMENT OF ENERGY. referring to several sexual orientation-specif- No. 91-5393. ic justify factors letting Krc serve abroad; despite a sworn witness’ statement United States Court of Appeals, security personnel USIA told her that District of Columbia Circuit. “problem” Kre’s conduct, was not his but his orientation; sexual despite an acknowledge- Argued May ment key USIA decisionmaker that he homosexuality considers per se to have seri- July Decided and, ous security consequences; finally, de- spite prior panel very decision this ease

clearly resting explicit on the ground that

Kre’s termination was status- not conduct-

based. All of these factors compel the con-

clusion that there is a serious dispute factual

here that deserves to be aired at trial and

that makes summary judgment wholly inap-

propriate. case, emphasize,

This I not a run-of-the-

mill fact controversy type in which we

routinely refuse en banc review. The affir- summary

mance of judgment in a case like implications

this has wide way for the this

court intends deal with recurring issue alleged government unconstitutional dis- against

crimination homosexuals —and for ability

our deal with similar mixed-motive

Title VII cases. The decision raises serious

questions well summary judg- our jurisprudence

ment in cases where the mov-

ant bears proof the burden of on the disposi-

tive issue. DC, argued Brown, Washington, W.

David cause, appellants. for Atty., Wash- Rees, Asst. U.S. S. Thomas cause, appellee. DC, argued the ington, Ramsey John- J. were brief him the With Craig R. Bates and D. Atty., son, John U.S. Washington, Attys., Lawrence, U.S. Asst. DC. GINSBURG, BADER RUTH Before: SENTELLE, Circuit

WILLIAMS Judges. by Circuit filed the Court

Opinion WILLIAMS. F. Judge STEPHEN Judge by Circuit filed Concurring opinion GINSBURG. BADER RUTH Judge: WILLIAMS, Circuit F. STEPHEN in a water sits heated When i.e., heat, environment, it loses a cooler law the second exemplifying thus energy, authority granted thermodynamics. Under Standards Conservation Energy by the Buildings Act”), (“Buildings New Act of 1976 percent per hour for oil gas water heat- (1988), §§ Department U.S.C. 6831-40 See 52 Fed.Reg. (1987). ers.2 Energy promulgated “standby addition, loss” rules DOE proposed to restrict the defi- to limit these losses in water heaters install nition of non-storage water heaters so that ed in new projects. federal construction some formerly heaters considered non-stor- Appliance Gas Manufacturers Association age would subject become to the standards. (“GAMA”)and five of its challenge members These revisions followed a proposal by the Department’s rule, 1990 version of the American Society Heating, Refrigerating claiming requirements are not in accord and Air-Conditioning Engineers, (“ASH- Inc. statutory with the authority RAE”) and are arbi to make its standard, pre-existing trary capricious. and Applying, a concept ASHRAE 90.1-1980, Standard more strin- developed from statutory different language gent. See ASHRAE Standard 90.1-1989 occupational context of (1989). (The health regula ASHRAE standards are devel- tion, the district court held that oped by rule industry experts; the nature of their conformed statutory to the mandate because record.) enforcement is by not revealed this it did not to, “threaten massive dislocation or Although argued GAMA the stricter imperil of, industry”. existence Gas proposal 1987 satisfy could not Appliance v. Depart standards, Ass’n adopted it Manufacturers as final in 1989. ment Energy, F.Supp. 773 466 GAMA then filed the first of two suits (D.D.C.1991) (“GAMA II”) (internal quota against DOE in the United States district omitted). tions and citations Because the court. Appliance Gas Manufacturers district court arrived at an incorrect statuto Ass’n, Inc. Secretary Energy, 722 ry interpretation and DOE apply failed to (D.D.C.1989) (“GAMA F.Supp. 798 I”), interpretation the correct adequately, re we the court set aside, the standards ruling that verse and remand to the district court with neither ASHRAE nor DOE had considered instructions to remand to DOE. Buildings requirement Act’s reg “ Background ‘practicable’ ulations be justifiable and Buildings terms of requires Act economic cost and Secretary' benefit im and pact Energy develop to affected energy groups.” interim Id. (quoting at 797 efficien cy 6831(b)(2)(1988) § standards U.S.C. that are to be binding for and drawing new 6839). § government federal It enjoined projects construction enforcement' of but merely guidelines rules and remanded the (though others case for the-oth DOE to develop reasons, ers include a statement of regulators). 'state and local to allow for See 6833(a), 6835, comment parties, §§ U.S.C. interested 6839. DOE and is to then to assure that issue its final response the standards “are rule and to adequately comments. Id. analyzed in at 798. energy efficiency, terms stimu nondepletable lation of use of sources of en remand, published On “Preliminary ergy, resources, institutional habitability, Adoption Statement Reasons for of Stand- benefit, economic cost and impact upon and by Loss (“Preliminary Statement”), Criteria” groups.” affected U.S.C. § 49,724 Fed.Reg. (1989), corrected, as In 1987 proposed revisions Fed.Reg. 50,341 (1989), to its attempted standby loss standards for commercial enjoined. stor- defend the standards GAMA criti- 1age heaters, cutting the maximum cized the Preliminary Statement as inade- (watts standby allowable W/sq loss to 1.9 ft quate proposed Although alternatives. per square area) foot of tank surface rejected for DOE proposal appellants’ kept electric water heaters and to 1.3- + the initial gas heaters, standard for oil 38/V "Storage” tank, water heaters gallon store hot water for the standard per- limit use, later "non-storage” whereas or “instantane- centage per loss Preliminary hour to 1.68%. See ous” heaters heat water as needed. Statemént of Adoption Standby Reasons of. Criteria, 49,724, 49,731-32, Fed.Reg. Loss 2. The formula percentage measures of heat loss 49,438 (1989). per "(X Y/V),” (expressed hour + where "V” tank). rated volume Thus for a 100- be issued standards to the inapplicable heaters electric requirements relaxed 6(b)(5) Whereas Buildings Act. under W/sq ft. 3.0 W/sq ft 1.9 from in- health human Rule”). protection concerns (“Revised Build- quantify, difficult risks volves lawsuit, present responded re- energy protection ings Act concerns deci reviewing the II. Rather consumers, costs involves sources statutory reference light sion easy quantify, relatively ’ benefits ap benefits, court costs economic computed. they be asks explicitly Steel of United two-step standard plied for standards calls Act Thus, the OSH while v. Mar America, AFL-CIO-CLC workers of from safeguard workers enough to stringent (D.C.Cir.1980), 1189, 1272 shall, F.2d 647 “extent workplace to the the in risks health were standards that determining first re- Act instead id., Buildings feasible”, that, be feasible, second and technologically “achieve to set be standards that quires was so market federal cause in ener- improvements practicable maximum adversely affect not could small, standard (emphasis § 6831 efficiency”, 42 U.S.C. gy industry. 'part of a substantial added). upheld The court 465-66. II, F.Supp. at 773 authority statutory Congress implicitly, within and explicitly rule as Both appealed. rules of explained. practicability adequately and that clear made entails Buildings Act under promulgated two appeal, of pendency During the ex- Congress declared First, cost-benefit ASHRAE occurred. events relevant “ade- be must standards such by that brought plicitly case of a settlement in agreed, ... economic of Ass’n, in terms analyzed quately GAMA, Appliance Gas Manufacturers affected upon benefit, impact and Refriger- cost Heating, Soc’y American v. Inc. Act’s § The' Inc., 42 U.S.C. Eng’rs, groups.” Air-Conditioning ating and efficiency and economic emphasizes 1989), 11, to preamble (D.D.C. Dec. 3319 89 Civ. No. concerned was Congress that 90.1-1989, clear adopt- makes Standard ASHRAE revise on energy waste impact the economic standard standby loss relaxed ing a new those pocketbooks heaters, consumers’ ASHRAE —and commercial all suppliers of funds ultimate as the Second, taxpayers (1992). Con- 90.1b-1992 Standard failure “[T]he government. by the loaned ASHRAE adopted the new gress mea- energy conservation adequate provide after manufactured equipment all such increas- buildings newly 1992, constructed sures Policy Act Energy 1, 1994. January may affect costs 2776, operating long-term es 122(d), 102-486, 106 Stat. § No. Pub.L. for, security of, and 6313(a)(5) (West adversely repayment § (42 (1992) U.S.C.A. by Fed- insured, guaranteed made, or loans construction For future Supp.1993)). federal instrumentali- regulated ... or DOE, agencies within eral directed Act projects, the ties_” 6831(a)(3). Congress § 42 U.S.C. 24, (on October enactment after years two energy costs thus concerned or ex was “meet 1992), adopt standards mortga- ability of reducing the 90.1-1989, were waste Standard of ASHRAE those ceed” This payments. mortgage them to meet gors than less no take effect standards new served hardly be well would concern Id., 106 Stat. financial § issuance. year a after mortgagors $100 saved 6834(a)(1)-(2) by standard § a (42 U.S.C.A. 2784-85 capital cost energy bills at annual leaves statute (West Supp.1993)). payments $120 mortgage their sup they are raised until undisturbed prior rules year. (presumably ones new by the planted (42 U.S.C.A. 1995). 106 Stat. October fac- refers § 6839 42 U.S.C. Of course moot. is not 6834(d)). issue Thus the benefit”; § cost “economic other tors efficiency, Interpretation stimulation “energy Statutory include energy, sources nondepletable of use expressed criteria substantive ... and resources, habitability, institutional construction as emerged Steelworkers might be There groups.” affected impact on Safety and 6(b)(5) Occupational § one allow statute where cases 655(b)(5), Act, 29 U.S.C. Health several of the trump other factors to review under that standard. And we need analysis, justifying adoption cosVbenefit of a remand the district court to apply the exceeding standard with costs standard, benefits. But correct may as we review the ad perhaps “impact not: groups” affected ministrative directly. record “We do not ac sounds like a stringency, constraint on any particular cord deference to the decision if, least as DOE sup district court of the where, District here, Court ‘the posed, Congress prevent aimed to undue dis District Court and this court are both re ruption industry. Appellee’s Brief viewing record, administrative and no ad 9; II, at 19 n. GAMA 773 F.Supp. at 466 & ditional testimony was taken in the District factors, n. 4. Three other “stimulation of use Court....’” Agency Costello v. Int’l nondepletable energy,” Den, sources of “institu (D.C.Cir.1988) F.2d n. resources,” tional “habitability,” are sub (quoting Walter O. Boswell Hosp. Memorial ject vague, speculative analysis Heckler, at best (D.C.Cir. 749 F.2d 790 n. 2 (and “habitability” 1984)). would seem to cap Cook, See also Novicki v. 946 F.2d benefits). tured in the costs Only “ener gy efficiency” and “economic cost bene susceptible

fit” are comparatively precise proceeding Before review, to that we two, Of “energy these note a efficiency” special number of First, issues. the suggests parties no itself, disputed reasonable limits and of have whether the statute may since it requires be increased to the limits DOE to of demonstrate a “path to technology long compliance”. so It nonenergy concededly resources does not do inso regarded are explicit as free. limiting princi As the concept terms. But the developed in ple inherent the occupational “economic cost context of benefit” and is health rules un both relatively determinate, 655(b)(5), § reasonable and der 29 U.S.C. where agency the any override of a need negative show analy little more than that .cost/beriefit the stan sis require would seem to very dards are technically careful feasible and that the . justification industry will survive the compliance. costs of context, Even in expressed that we wonder any event, although DOE’s brief at how OSHA “could general arrive at even impact stressed the groups, affected estimates compliance of costs when is so Department not, did at the time rule- unclear employers as to how comply”, making, consider why or whether that factor Steelworkers, 1296, 647 F.2d at and we de (or any other justi criterion of would manded a clear explanation, id. The need fy adoption of a marginal standard whose for a discernible surely is costs exceed its marginal benefits.3 Thus we more compelling agency where per .must (1) may say that Congress clearly deter form a direct balance of costs and benefits. mined that perform DOE must a proper Inc., analysis, cosVbenefit Chevron U.S.A. v. Second, plaintiffs object that NRDC, 837, 842-43, U.S. 104 S.Ct. agency unduly computer relied model (1984), 81 L.Ed.2d 694 that in ing. Our approach to fairly this is well es present ease DOE has so far identified tablished. We have although noted that no for disregarding bases the outcome of “computer modeling is a useful and often analysis,to such an impose a standard that essential tool for performing the Herculean fails that test. labors Congress impose[s] on [administrative agencies]”, models, despite “[s]uch com their Application the Statutory Standard plex design and aura of validity, scientific are purported As DOE apply at best imperfect subject cost/ben to manipu efit analysis I, on the remand from GAMA Costle, lation”. Sierra Club v. 657 F.2d we have before us susceptible a record accuracy Since the 3. We justifications must consider those subsequent that DOE's endorsement of "massive DOE used at the rulemaking, time see SEC dislocation test” used the district court in 80, 87, Chenery, 454, 459, II, 318 U.S. S.Ct. F.Supp. Appellee's see (1943); L.Ed. 626 therefore we do not consider Brief at 19 9.n. de- course we Of supported. inferences on whether “hinges model computer technical id., scientific any relevant reality,” fer to reflect assumptions underlying tous authorize not does but expertise, explain the sufficiently must agency reason- of DOE’s steps the critical gloss over pre- methodology used assumptions ing process. a com- model; provide it must paring its model defense analytical plete Analysis DOE’s Cost/Benefit reasoned objection with each respond Heaters complexity Water A. Electric technical presentation. agency relieve not analysis does Appellants argue that DOE’selectric factors relevant all to consider burden standard standby loss heater storage water to its stones stepping identify the capricious. arbitrary and W/sq ft of 3.0 a rational must There decision. final such no claim Although appellants inputs, factual between connection can meet market currently on modeling results assumptions, modeling true, not invali- fact, standard,4 if results. from drawn conclusions could demonstrate if DOE date footnotes, internal (brackets, 657 F.2d at costs the standard attain firms could omitted). The marks, citations quotation That is § 6839. demands satisfying the when heavier becomes burden agency’s down. has fallen where on to being relied “is prediction method ob Factor. GAMA Fudge 1. The 40% test data...." actual ... ‘adverse’ overcome faulty modeling is computer jects Ruckelshaus, v.Co. Harvester International results correspond to the does (D.C.Cir.1973). As we 615, 642 F.2d ' theOn DOE directed. testing that limited here. the case see, was that shall See is valid. tests, objection the face par argument that DOE’s note Finally, we 23,853-54. Fed.Reg. at Rule, Revised 55 “is agency an when is due ticular deference heat predicted had DOE model the Whereas special of its area within making predictions, be heater electric for modified loss a Balti of science.” frontiers the expertise, at 294.2 was Btu/h, loss Btu/h. actual 193.4 Re v. Natural Co. Electric Gas more predicted model computer whereas the And 103, 87, Council, 462 U.S. sources Defense R- in an covered unit same for the loss heat (1983). 2255, 437 2246, L.Ed.2d 103 S.Ct. batt blanket fiberglass insulation is case reality “this that respond Plaintiffs ft),5 (or the actual W/sq 2.19 162.3 be Btu/h improvement incremental hypothesized about ft),6 in (or W/sq 3.21 237.9 loss was Btu/h well- conventional exceedingly responded standard. of DOE’s excess technology, transfer heat of field plowed im marginal on the primarily a focus with ener increased any new ‘solution’ where batt, yielded which adding the R-ll of pact easily tested readily and efficiency can gy by the model predicted than savings greater equip existing state-of-the-art modifying Btu/h, a 25.2 31.1 (56.3 father Btu/h the ‘solution’ with in accordance ment difference). was model Since Btu/h transfer heat unit’s measuring simply in savings, DOE predict jacket intended 16-17. Brief at Appellants’ performance.” in sav 25.2 additional Btu/h ferred compliance about fact, expectations In reducing heat by the batt’s ings caused was daring any not from derive its standards This Maybe so. controls. from loss from physics, but particle into insight new to solve however, nothing did theory, proposi accepted applications purported To heat losses. high absolute problem simple rather is whether issue tions. highlight deficiencies doing would so whether doubt appears be some 4. There and esti- factor Using conversion DOE's model. 120-gallon electric particular one tank, 50-gallon area of the surface mate of Appellee’s standard. meet could model at Conference for Pretrial Motion Opposition to 8; Wenczl Ernest Declaration n. at 16 Brief (May 162.3 / as follows: Btu/h calculation 12; 23, 1991) Declaration Second at = W/sq ft. Btu/h/W) (21.7 sqft)] 2.19 x [(3.412 (June at 2. Wenczl Ernest ft)] (21.7 Btu/h/W) sq x [(3.412 figure 6. 237.9 Btu/h into / convert not itself Btu/h did 5. DOE = W/sq ft. standard, 3.21 because possibly W/sq ft of the explain away, DOE assumed them to rely on its own expertise to resolve fittings, be due losses from issues, had not doubtful factual but that “may it during modified been the tests. DOE made tolerate needless uncertainties its central assumption fittings losses from assumptions when the fairly evidence allows equal 23,- Fed.Reg. 69.9 Btu/h. investigation and solution of those uncertain nearly up 854. This made the otherwise ties.” Council, Natural Resources Defense unexplained between the model difference Herrington, Inc. v. 768 F.2d o (to prediction and the test results within 5.7 (D.C.Cir.1985); see als National Lime Btu/h). EPA, (D.C.Cir. Ass’n v. 627 F.2d 1980). important, An easily hypoth testable model, computer its.revised DOE then esis should not remain untested. Where cre postulated 40% reduction in at the .losses fully ation of a complying prototype fittings controls, is not spec- but nowhere did it very costly, achieved, appears to be the ify case for how these reductions are to be here, water heaters it seems percentage This reasonable to do incorporated was into the so in order to validity establish the model and has subsequent predic- affected all energy-saving benefits (Table 8). agency be tions. .See lieves can be achieved at a Following model, this reasonable cost. specifies an addi- Thus, DOE should consider the of 3.5 costs and tion inches of fiberglass 75% foam/25% testing benefits of options, taking insulation into ac as a 50- count importance hypothesis, gallon water heater its purposes for the uncertainty, the likelihood that testing would Statement of resolve Reasons, uncertainty, and the cost of Tab 7. test ing. While of course we would defer to Defending handiwork, its DOE asserts reasoned decision on testing, incremental estimates of losses fittings due to here we cannot discern even the faintest controls are accurate and were derived from glimmer of an effort to make such a decision. various properly studies repu deemed an Arbitrary Use Multiplier table. 55 Fed.Reg. at Let us as Projecting so, Key Costs. Department’s to the sume and also that DOE is correct projection's cost unexplained were assuming that as- problem there was no in ap sumption that the costs including plying the insula- studies - which were based oh residential tion in a residential electric water heat heater will commercial ers- to ones. Even typically be 2.5 times the costs of assumptions, these the insula- figures with the do not (2) tion itself and an assumption support that the assumption DOE’s first that 40% of these assumption could be extended to the be com- losses eliminated at a cost. reasonable mercial market. vigorously DOE cannot Plaintiffs accurately resist estimate the costs step, the adducing figures second compliance and benefits of indicating without some no that the residential industry tion how market is will achieve 170 times the the loss size of Thus, reduction. may they The commercial market. complex; task be say, fittings vary the unit according product costs of redesign to the model retooling higher manufacturer. 55 are far at Yet diminutive some method of compliance commercial market. pro They must also claim unit in posed production order to provide legitimate higher, costs are relatively foun due to for low-speed, dation production' cost-benefit labor-intensive lines. 28,1990) See (February Comments at Testing. object Plaintiffs to 15-16. DOE’s failure pursue to testing its to the (cid:127) point having on hand at least one real- in difference total size of the two world electric water heater that may satisfied the necessarily markets not undermine applied standard. We have a common sense DOE’s extension of multiplier the 2.5 to the matters, test to such saying agency that an that, commercial market. It is conceivable 23,853-54. 7. DOE added 69.9 fittings Reg. Btu for the inferred at Note fact the 25.2 losses, losses and 25.2 for inferred controls for a controls, supposedly savings was pre- at the not 95.1, total of or less than 5.7 Btu 100.8 losses, savings point but DOE’s error on this difference computer between what the model plaintiffs. favored predicted prototype and the achieved. 55 Fed. 8 temp- a smaller utilized analysis calculations difference, firms aggregate despite the mod- effect ... which differential erature same might be industries in both

plants analysis.” resulting cost benefit erated characteris production have similar size Confer- Pretrial for to Motion Opposition See failed conspicuously has so, if DOE But tics. added). “Moderat- ence, etc., (emphasis at 4 narrowly, to or, more make to ease - supposed exaggerated simply It ed”? contentions plausible the focused answer In its standard. from DOE’s energy saving Fabri See, e.g., International plaintiffs. energy calculated analysis, DOE cost/benefit 384, 389 EPA, F.2d Institute care ft) (2.47 with W/sq be 122.21 loss to Btu/h report ato points DOE fiberglass insulation. 75% 3.5-inch foam/25% evidence Inc. as Little D. Arthur with recognized that later Although DOE impact no have differences size market itself the the standard used differential is report at but Brief cost, Appellee’s id., ft, W/sq 2.77 been have loss would residential to the expressly directed using benefits to recalculate no effort made Efficiency Products market, Consumer see leaving benefit thus figure, this corrected Document, Analysis Engineering Standards analysis component eost/benefit D3-1, ap Little Inc. D. by Arthur profit figures addition, as both skewed. to address differences pears nowhere factor, fact that mere fudge 40% from the of its Because markets. the two between ft) (2.77 W/sq figure corrected ac critiques, DOE’s answer failure of a no evidence comply with the decisionmak- reasoned manifest not did tion cannot compliance. path to reasonable arbitrary and that extent ing and was the standard for of conditions one set use capricious. set, to another, more favorable itself, and method’s compliance proposed estimate argues claim, GAMA related In a purposes. likely achievements com- its addressed adequately not costs. shipping increased about ments Heaters Gas Water B. Oil (February Comments justify a failed again,9 Here seems objection Standing alone GAMA’s its demonstrate shipping weak; contentions rather costs. their exceeded benefits standard’s comparison make no heaters of commercial (1) “high-input” paths proposed appear ones, do thus residential with of foam of two inches heaters,10 the addition Depend- multiplier. 2.5 inconsistent insula- fiberglass inches or three insulation however, remand, course ing upon the heaters,11 the “low-input” tion and pertinent. may become the issue installation and the of insulation addition points nei- IIDs. DOE dampers and flue Temperature Incorrect 4. Use Dif- prototype nor a model current ther a Analysis. Rather in Cost/Benefit ferential the standard. meets n. at 23 Brief Appellants’ obscurely, see electric a lower from the used familiar problems DOE objects that Two require no repeated en- and are to estimate context differential” “temperature First, again ex- analysis DOE discussion. its further savings in ergy the residential multiplier from itself the 2.5 tended the standard (67.7°F) used in Preliminary market, see acknowledged commercial to the ultimately (80°F).8 itself DOE 49,740-41, 49,733, Statement, Fed.Reg. at “cost benefit saying its discrepancy, ratings input have "High-input” water heaters temperature 10. between the difference 8. This is more, typically are 155,000 al- or and surrounding room Btu/h and water of the stored temperature. differential, dampers "IIDs" and greater equipped flue ready with The Statement, Preliminary devices). ignite heat. The latter (intermittent ignition loss of faster the 49,732. combustion, Fed.Reg. at rather than required to only start as lights. pilot continuously do burning as gas water oil and objections for treats 9. GAMA to concen- has chosen the same as heaters trate on ratings input heaters have "Low-input” water Although DOE 11. gas heaters. water Btu/h, 155,000 argu- any thereby forfeited between that GAMA claims heaters, acknowledge dampers IIDs. flue equipped we water with oil are ments 10% together. them will treat similarities extension, justifying second, without requiring ed those Congress devices. did not justify it failed to delegate absence of additional the standard-writing ASHRAE, testing. problems real-world The other nor authorize DOE to do so. Thus DOE’s gas distinctive to the regulation oil heater market. on its own. must.stand Disregarding 1. Adding the Cost Flue As manufacturers have not found it worth- Dampers Low-Input and IIDs to Heaters. while to add these low-input devices to heat- Low-input water heaters significant constitute a ers independently of the standby-loss re- por the market for com quirement, it is standby-loss concern that tion- 2 5%—of gas mercial oil and water heaters. Second will cause the installing cost of them to be Joseph (if Declaration of Mattingly (May M. incurred indeed ultimately adopts 1991) agree at 7. All that insulation them). alone requires Accordingly, will not enable these to meet the standard the cost must be-included if the cost-benefit installing and that flue dampers analysis and IIDs is is to be a coherent analy- marginal (To necessary. not, however, DOE has sis. included the extent that the devices contrib- d any installation, cost for such an its ute to the value of the water apart heaters analysis in Preliminary standby-loss reduction, from cosi/benefit that windfall Statement did not even mention those costs. taken-into account as an offset to the cost.) protested these modifications would lead to price' increase $500 Computer The Model. As electric more, represents some cases 50% of heaters, DOE encountered inconven- price. current unit Appellants’ Brief at 32 & ient facts testing limited gas 30; (Feb. 1990) n. GAMA’sComments at attempted heaters and to brush off the flaws responded by claiming that since thus revealed in its computer model.' When equally stringent ASHRAE standards would the National Institute of Standards and soon be effect the entire commercial oil (“NIST”) Technology 78-gallon tested a sin- gas market, water heater require DOE’s gle-flue gas water heater wrapped in an R-ll impose ment would no additional' burden: fiberglass blanket, insulation measured “Thus the added cost of these devices is standby hour, per losses were op- 2.37% virtually they since nil have been re posed per to the required 1.76% hour by the quired anyway in the absence of the DOE standard. Fanney Letter from A. to DOE standard,” Rule, Revised (May response, at 2-3.13 In 23,863. compliance recommended for “the addition of one fiberglass inch of to the base two fact, inches In predictions regarding in [the R-ll or ... approxi- blanket] the use of dustry proved false, standards to be as ASH- mately two inches of foam insulation.”- adopted RAE stringent less standards for Fed.Reg. at GAMA attacks DOE’s low-input models12 —standards with which computer grounds model on several and con- they can evidently comply without flue dam tends that proffered path (or pers or appellants IIDs so tell us without proven is not to be effective. contradiction, Appellants’ see Brief at 33 n. 32). event, DOE had no basis to a. percentage through jacket. loss assess the costs new simply devices zero percentage through of loss the heater’s because regulatory body contemplat- jacket another significant because the main 12. input ratings however, standard for units company emphasized, with of that it was 155,000 slightly is less than Btu/h strict that for testing conditions, under somewhat unrealistic higher-input Compare units. 42 U.S.C. including external insulation around controls 6313(a)(5)(B) ((1.30 114/V)%) + with fittings possible that would not be for a unit 6313(a)(5)(C)((1.30 95/V)%), § § by + as amended West, operation. in actual E. Branford White’s 122(d) Energy Policy of the Act of 106 (Feb. 28, 1990) Although Comments at 3. DOE Stat. 2812-13. provided by claims that "[t]est data the Califor- that, Energy nia Commission indicates several Branford-White, a manufacturer and member gas water commercially that are heaters avail- GAMA, of performed testing its own and arrived standard],” comply able [gas do with the slightly at a value closer to the standard than the 23863/1, Fed.Reg. at no evidence in the record testing NIST standby had. Branford-White found a supports this assertion. hour, per of compared loss 2.02% requirement to a per 1.68% hour under standard. (at study that aon based in turn was jacket outlines - increased DOE D.E. Slaughter and forthcoming), G.G. time losses. only these affect insulation - would Con- Evaluation Efficiency An Spann, accounts jacket loss that asserts DOE Wa- Domestic Analysis Economic sumer gas water standby loss of total of the 70% (1978) at 19. Heaters, Con-5 IIDs, ORNL/ 55 Fed. ter dampers and flue with heaters heat- water residential out that points that the asserts 23,857, while Reg. at dampers, 50%, flue have do not typically approximately ers is figure corresponding light pilot in lower results (Feb. at 8-9. that lack that Comments GAMA’s study points out Indeed, the 1978 efficiency. anything us to pointed DOE restricted, as it severely of is assumption 70% if a flue that supports that record effi- light damper, pilot objec by a flue to GAMA’s response In jacket loss.14 Thus at 160°F. Id. is ... conceiva to 71% ciency “it rises that tions, declared DOE idiosyncratic and the its rather production to relate effort design DOE’s better that ble on appears based dam flue world the real of to prototype use electronic predominant more heat the residual misconception. much of rectify a pers 23,858. Rule, Fed.Reg. at Revised loss.” Storage Units Redefining is what C. to speculation as to DOE’s resort bank simply illustrates adequate “conceivable” an perform to failed also DOE to path a delineate its effort analysis, of ruptcy any analysis, or indeed cost/benefit compliance. After storage units. its redefinition as storage heaters defining water initially proto- test atypical an Adjustment b. 4,000 rating less than input an with water those high-input ,gas most Unlike type. (cid:127) storage capacity a gallon and per heater DOE water Btu/h heaters, 78-gallon Preliminary State- model, gallons, ten more than to correct tested, was used which ultimately 49,727, DOE ment, Fed.Reg. at Instead IID. not an light pilot had rating of input with an attempted as those them model, defined DOE ordering a different storage per gallon adding a by less Btu/h peculiarity adjust for this “ Fed. gallons. 55 ten ‘Q more than explained, capacity of As factor, ADD.” DOE “Q so, adopted doing DOE must be Reg. at amount losses is the ADD’ broadened total standard ... to match jacket an ASHRAE losses added to include storage water heaters loss- measured [testing agency] definition with losses 23,877/3. instanta- formerly considered Fed.Reg. Rule, at that were some Revised es.” 90.1-1989, printed in ASHRAE apparently underestimated neous. Because DOE claimed 23,865/1. circum- DOE light such at pilot Fed.Reg. efficiency of a consciously estab- of re- benefit “was stances, new standard overestimated it over the light with confusion pilot prototype’s to eliminate placing the lished in- heaters in turn led water storage IID. This definition in a loss nor- non-jacket heat heaters.” stantaneous underestimate (one however, no appears, gas water heater 23,865/1. far high-input mal So i.e., loss, simply jacket surfaced; was IID) overstate thus to had confusion —and DOE’s impor- loss addressed More only kind standards. following ASHRAE high-input eco- compliance for standard’s analysis proposed tantly, specific Although lacking. entirely heaters. is effects nomic anal- “the here asserts from, a stem difficulties again, Once for the valid performed ysis that DOE the residential analogy between dubious cov- heaters storage water range entire light put pilot markets. DOE commercial criteria”, Appellee’s standby loss by the ered a 1977 22%, figure based efficiency at supporting nothing points Brief heaters, R. Hos- residential study for figures— contention, its own indeed Labora- Hirst, Ridge National E. Oak &kins very heaters— small address don’t Analysis Residen- Energy and Cost tory, declining gradually ratios 33, (June show benefit/cost Heaters Water tial unsupported appear again figures although Statement, assumed Preliminary In the

14. 49,- Fed.Reg. at record. by anything in the depending on the percentages varying jacket loss IIDs, dampers and 733/1. of flue presence or absence See, Statement, e.g., Preliminary § size. However, listed 6839. Congress consid- meaningful, goal To be ered the averting energy an shortage significant analysis enough DOE’s economic must take into ac- merit mention in its statement findings differing purpose. count the standards’ effects on the § n 6831(a)(2) (“[Standards major segments newly regulated. the market con- buildings structed prevent ... waste of energy, which the Nation longer can no af- substantially defects identified under- ford in view of its anticipated current and Department’s justification mine the of the ' energy shortage.”). Referring to the use of Revised Rule. To the extent that a “feasibility” context, test in the OSHA properly15 raised explicitly other claims not court observes protection that the of human here, persuaded addressed it has not us that health involves losses quantify. difficult to they had substantial merit. This is not to depletable Protection of energy sources of however, deny, polished that more versions also involves losses quantify, difficult to critiques of these raised the course of the appears therefore to warrant a test less ex- may require remand to DOE consideration acting than a by DOE. preliminary injunction Whether should against

issue part enforcement of all or

rule, compare F.Supp. GAMA 722 at 797-

98, rests within the sound discretion of the court, principles district under the governing MOVING PHONES PARTNERSHIP agency interim relief an when rule is found L.P., Appellant, See, Union, wanting. e.g., International v. FMSHA, (D.C.Cir. 960, UAW v. 920 F.2d FEDERAL COMMUNICATIONS 1990) (pointing to “the seriousness of the COMMISSION, Appellee, (and order’s deficiencies thus the extent of agency doubt whether the Thomas correctly) chose Domencich and Committee for a Lottery, disruptive Fair consequences Engineering, and the of an Romulus inter Inc., Martin, Jr., change Cellular, im James E. may CSH changed”); itself be Al Cellular, L.P., Cellular, Inc., Gilcom PC Signal, lied Inc. v. U.S. Regulatory Nuclear Company, RSA Cellular AAT Comm’n, R A 146, (D.C.Cir.1993). S 988 F.2d Company, L.P., II, Excellence Elleron Accordingly, we reverse and remand the case Cellular Sunde Cellular Corporation, the district court for preliminary relief Communications, Inc., Intervenors. may appropriate and to remand to 91-1611, DOE for further proceedings. through 91-1617, Nos. 91-1613 91-1621, 91-1626, 91-1627, 91-1628, So ordered. 91-1629 and 91-1636. GINSBURG, RUTH BADER Circuit United Appeals, States Court of Judge, concurring: District of Columbia Circuit. I concur in opinion the court’s with one 8, Argued April 1993. Assuming, states, caveat. as the court July Decided the “stimulation of use nondepletable re- Rehearing Suggestions for Rehearing sources” factor in 6839 is more difficult to En Banc Denied Oct. apply than the “economic cost and benefit” factor, it does not necessarily follow that

DOE must discount the former factor. It is

true that this factor is one of several NLRB, 15. We do not properly (D.C.Cir.1990) (“We consider claims not 899 F.2d 50 n. raised, argument Appellants’ such Reply require petitioners appellants to raise all of Brief at unjustifiably long 18 that DOE arguments opening prevent their used brief to (a "payback period” concept evidently ‘sandbagging’ appellees respondents used all hands as a crude provide substitute for the opposing more counsel the chance to re- discounting INS, conventional spond.”); Reyes-Arias costs and benefits to see also 866 F.2d value). present See Corson & Gruman Co. v. 504 n. 2

Case Details

Case Name: Gas Appliance Manufacturers Association, Inc. v. Department of Energy

Court Name: Court of Appeals for the D.C. Circuit

Date Published: Jul 27, 1993

Citation: 998 F.2d 1041

Docket Number: 91-5393

Court Abbreviation: D.C. Cir.