“SECRETARY OF LABOR,Complainant,v.SEIBEL MODERN MANUFACTURING &WELDING CORPORATION,Respondent.OSHRC Docket No. 88-0821DECISIONBefore: FOULKE, Chairman; MONTOYA and WISEMAN, Commissioners.BY THE COMMISSION:At issue is a single serious citation item alleging that \”[h]elpers orattendants\” did not have \”proper\” eye protection while performing tackwelding operations.[[1\/]] The administrative law judge vacated thecitation item on the basis that the helpers involved in this case werenot exposed to hazards necessitating eye protection. The judge did notreach other issues, including the employer’s affirmative defenses. Forthe reasons that follow, we reverse the judge’s decision and remand thecase for a determination consistent with the principles of law discussedherein.I. _The Employer’s Tack Welding Operation_During several months in late 1987 and early 1988, Gordon DeLeys, acompliance officer with the Occupational Safety and HealthAdministration (\”OSHA\”), of the United States Department of Labor,conducted an inspection of Seibel Modern Manufacturing & WeldingCorporation (\”Seibel\” or \”the employer\”) [[2\/]] Seibel manufacturesindustrial-size, metal battery cases at a fabrication shop in Lancaster,New York. On five separate occasions at the shop during his inspection,Compliance Officer DeLeys observed a tack welding operation performed bytwo welders with two helpers, working in pairs.[[3\/]] Each pair ofemployees, a welder and a helper, were fitting together, and tackwelding into position, metal pieces of battery cases. Tack welding holdsthe metal pieces precisely in place for the next step in themanufacturing process, continuous welding.Seibel has two tack welding operations, a one-employee operation forsmall battery cases (30-50 pounds), and the two-employee operation atissue here, handling large battery cases (up to 1,000 pounds or more).In the two-employee operation, which Compliance Officer DeLeys observed,the welders were wearing welding helmets having appropriately shadedlenses. However, the helpers eye protection consisted solely of, atbest, street-wear prescription spectacles having clear (unshaded)lenses. Because the welders and helpers were the same distance(approximately two feet) from the arcs, Compliance Officer DeLeysbelieved that the helpers, like the welders, should be equipped withappropriately shaded lenses. The risk created by the absence of thisprotection was welding flash–that is, corneal burns which, if repeated,could lead to blindness.Seibel’s president, Leon Seibel, who testified at the hearing, was \”notpersonally aware of any cases of weld[ing] flash that either of my two[helpers] have indicated to me.\” One helper had worked in the tackwelding operation for eight years; and the other had worked fortwenty-five years. President Seibel suggested that helpers did not needshaded lenses because the metal pieces being welded often shielded thearc from view and, in any instance where metal was not successfullyshielding an arc, the welder could usually cup a gloved hand around it,to block the helper’s view. Nevertheless, Compliance Officer DeLeysobserved the helpers looking toward bare arcs and, on the basis of theseobservations, estimated that the arcs were unshielded 25 percent of thetime. President Seibel acknowledged a possibility that the arcs were notalways shielded, and he posited that a helper could, as a last resort,look away from an arc. However, as we will describe in greater detail,the president’s testimony describing the operation tends to confirm thata helper generally needs to watch whether the metal pieces are fittingtogether correctly. Compliance Officer DeLeys testified along the samelines, that a helper frequently must look toward the arc to assure thathe is holding the metal pieces in the correct position.President Seibel testified that the helpers use rulers and must be ableto see well enough for precision within plus or minus 1\/32 inch.Therefore, according to the president and the employer’s expert witness,Gary Howard, of Mills Welding Supply, the helpers could not wear thesafety glasses or goggles with deep shaded lenses that were suggested byCompliance Officer DeLeys.[[4\/]] President Seibel particularly notedthat deep-shaded lenses would prevent a helper from seeing the rulermarkings while he is lining up the metal pieces and fitting themtogether. According to President Seibel, a welder wears a deep-shadedlens while making a tack weld only in order to be able to see duringthat short time: \”[T]he light [from the arc] is generally so bright that[a welder] would not be able to see what he [is] doing unless he ha[s] ashaded lens.\” But, President Seibel indicated, if a helper were requiredat all times to wear safety glasses or goggles having deep-shadedlenses, he would not be able to see during the process of assembling themetal pieces, which is done in ordinary light, in preparation for thetack welding.Compliance Officer DeLeys suggested that this difficulty could beeliminated by having the helpers wear welding helmets or flip downshades, and flip these temporary forms of eye protection down only whenready for a tack weld. President Seibel indicated, in reply, that ahelper cannot \”shake his head down similar to what a welder does\” (tocause a flip-down helmet to fall into place over the face). A helper’shands are occupied with keeping the large, often warped metal piecesfrom shifting, and the action of flipping a helmet into position couldcontribute to imprecision:Q. What, if anything, would prevent your [helper] from lining up thematerial with the ruler or measuring device, [then] when he had it linedup, putting on a dark shaded eye protection, and then communicating tothe tack welder, proceed?A. He would not be able to see what the material has done as soon as heput his eye shade on compared to when the tacking operation commences.Q. He has already lined it up, what is there for him to–A. It doesn’t–it doesn’t stay lined up as easily as you are alludingto….[I]t is essentially level material and when it is sheared, it inmost cases will get a slight bow to it and in doing so[,] when we putthe material together, certain force has to be exerted to spring thematerial into position and [to hold it] there at the instantaneous pointwhere it has to be tacked. And if he were to have a dark shade on, hewould not be able to see the ruler operation[:] the ruler dimension heis trying to [adhere to] as he is holding the material and trying tospring it in[to] position.The rulers used by Seibel were yellow with black markings. In responseto questioning by the Secretary’s attorney, Seibel’s president testifiedthat Seibel had not tried using \”magnified rulers.[[5\/]]Nor had Seibel tried using a jig, which is a mechanical device used tohold pieces of metal in the correct position while they are beingassembled. President Seibel testified, however, that a helper could notbe replaced with a jig because, for the largest battery cases, the jigwould have to be \”enormous.\” Also, a jig would need to be adjustable tofit many different-size cases, or there would need to be manydifferent-size jigs:[W]e have about 5,000 different sizes and we only run these batterycases sometimes one or two at a time[,] maybe five or ten at a time. Ajig operation is out of the question.Specifically:[W]e can not have a jig for every size. We get new sizes constantly, newdimensions constantly[;] the dimensions are [often] changed and it wouldbe strictly inappropriate to be able to jig up that particular materialand if we could, we would have a long time ago, believe me.Moreover:You can not make [a jig] adjustable for as many varied sizes and types.They are not all square steel boxes. Some are rectangular, some haveindentations in them, some have appurtenances that have to be added tothem[;] they have different types of hinging on them in some cases.Lifting apparatuses have to be attached to them. Without you knowing theoperation, it is extremely difficult for me to try to describe all thedifferent things that go on in that particular operation to a point in time.Q. Correct me if I am wrong. But what you are saying is not [that] itwould be impossible, but rather [that] it would be very time consumingand therefore very costly.A. No. I am saying it would be impossible.President Seibel went on to note that making adjustments to a jig on apiece-by-piece basis \”sounds very simple the way you are describing itand I wish it could be that simple, but it is not.\” The most substantialproblem would be that the jig could not properly hold in place the manyheavier metal pieces that are \”slightly bowed,\” therefore requiringpressure to force (or \”spring\”) them into the precise position.In addition, President Seibel suggested that there could be an increasedrisk of welding flash if helpers began to depend on flip-down eyeprotection. According to President Seibel’s testimony, a helper withoutany eye protection could shield himself behind the metal pieces and, ifthat proved ineffective, the welder could cup his gloved hand around thearc. However, if a helper had eye protection to put on, neither he northe welder would concentrate any longer on providing shielding; instead,the welder might proceed with the weld while the helper was still tryingto adjust his flip-down lenses or shake down his helmet. As PresidentSeibel stated: \”If a [helper] mistimes his helmet coming down to when hetells the [welder] to tack the two pieces of metal and my [welder] doesnot use a hand over the arc flash as they have alluded to, then we wouldhave more probability of having weld[ing] flash to my [helpers].\”II. _The Secretary’s Proof of a Violation_To establish a violation of a standard, the Secretary must show by apreponderance of the evidence that: (1) the cited standard applies, (2)its terms were not met, (3) employees had access to the violativecondition, and (4) the employer knew or could have known of it with theexercise of reasonable diligence. See, e.g., Walker Towing Corp., 14 BNAOSHC 2072, 2074, 1991 CCH OSHD ? 29,239, p. 39,157 (No. 87-1359, 1991).In this case, there is no question of applicability, as the citedstandard covers \”all arc welding\” operations and covers \”[h]elpers\”; noquestion of access, as the helpers could and did look toward the weldingarcs in the normal course of tack-welding work; and no question ofknowledge, as Seibel knew of the conditions under which the helpers wereworking. The one question relating to the Secretary’s burden of proof iswhether the terms of the standard were not met, that is, specifically,whether Seibel failed to comply with the cited standard’s requirementfor \”proper eye protection.\” Seibel contends that clear lenses weresufficient eye protection for the helpers, and therefore \”proper\” withinthe meaning of the standard, because of a relative lack of hazard fromthe welding operation as it was performed at Seibel’s fabrication shop.Former Administrative Law Judge Foster Furcolo agreed, concluding thatthe \”uncontradicted testimony that the Respondent’s method of weldinghad been carried on for some 25 years without any accident or injury toany employee\” shows \”that its method of welding is non[ ]hazardous.\” For the following reasons, we set aside the judge’s decision and holdthat the cited standard was violated in this case.The cited standard, 29 C.F.R. ? 1910.252(e)(2)(i)(a), governs the\”selection\” of \”[e]ye protection\” for employees involved in welding. Itexplicitly requires that \”[h]elmets or hand shields shall be used duringall arc welding operations\” and that \”[h]elpers or attendants shall beprovided with proper eye protection.\” See note 1 supra. The standarddoes not specify that the degree of protection \”proper\” for helpers isthe same as for welders, but in a nearby, related standard, 29 C.F.R. ?1910.252(e)(2)(ii), there are apparently comprehensive \”[s]pecificationsfor protectors\” that imply that lenses having a degree of shading arepart of \”proper\” eye protection. In other words, the specificationsindicate that clear lenses are not \”proper\” protection. Notably, two ofthe specifications are the following:Lenses shall bear some permanent distinctive marking by which the sourceand shade may be readily identified. The following is a guide for theselection of the proper shade numbers.These recommendations may be varied to suit the individual’s needs.[[6\/]]Also, the final specification of 1910.252(e)(2)(ii) states:All filter lenses…shall meet the test for transmission of radiantenergy prescribed in ANSI Z87.1-1968–American National StandardPractice for Occupational and Educational Eye and Face Protection.The significant piece of information to be gleaned from the referencedANSI test is that clear lenses are not \”filter lenses,\” and are notmeant to give the same type of protection. ANSI section 6.3.3.3.3 statesthat \”[c]lear lenses shall transmit not less than 89 percent of theincident luminous radiation\” (emphasis added). But ANSI section6.3.3.3.2 states that \”[f]ilter lenses shall meet theradiant-energy-transmission requirements specified in Table 1.\” Filterlenses, as defined by the latter ANSI section, are shade numbers 4through 14. Shade numbers 1.7 through 3 are \”[a]bsorptive lenses,\”according to section 6.3.3.3.1, which also specifies that this type oflens must meet the \”radiant-energy-transmission requirements\” of Table1. The table limits not only luminous transmissions but infrared andultraviolet. Actually, according to the table, even the lightest shadenumber–1.5 (which apparently is not even an absorptive lens)–cutsluminous transmittance to 55-67 percent, infrared to 25 percent, andultraviolet to 0.2 to 65 percent (different percentages for fourdifferent wavelengths). In sum, these lenses–filter, absorptive, andother shades–must meet specific requirements limiting transmittanceacross the full spectrum of light. Clear lenses, however, are notsubject to these requirements and actually must not fail to transmitmost of the luminous portion of the spectrum.On review, [[7]] Seibel maintains that the cited standard does not stateor imply that shaded lenses are \”proper\” eye protection in an arcwelding operation employing an electrode smaller than the ones listed inthe guide accompanying the cited standard, _see_ note 6 _supra_, andSeibel endorse the judge’s rationale that the Secretary failed to provea hazard. As we have discussed, however, and as the Secretary argues,the cited standard implicitly presumes a hazard where clear lenses,only, are being used. For every employee involved in arc welding,welders and helpers alike, the standards, requires a degree of eyeprotection that includes shaded lenses. _CF._ _Vanco Construction,Inc._, 11 BNA OSHC 1058, 1071, 1983-84 CCH OSHD ? 26,372, p. 33,454 (No.79-4945, 1982), _aff’d_, 723 F.2d 410 (5th Cir. 1984) (a broadly wordedeye protection standard is not vague because there is a \”guide in theselection of eye and face protection\” that informs employers ofappropriate protection for various operations). The fact that smallerelectrodes of the size used by Seibel are not listed in the guideassociated with the cited standard does not mean that helpers exposedto arcs from such electrodes may work without protection. The \”guide\”is exactly that, a guide, and employers must make reasonable effortsapply its guidance to their own operations. _Cf._, _Towne ConstructionCo._, 12 BNA OSHC 2185, 2188-89, 1986-87 CCH OSH ? 27,760, pp.36,310-11 (No. 83-1262, 1986), _aff’d_ 847 F.2d 1187 (6th Cir. 1988)(load chart listing some weight limits applicable to specifiedsituations provides sufficient notice to use an extrapolation method todetermine the precise limit in other situations).[[8\/]] Because thecited standard presumes a hazard where clear lenses rather than shadedlenses are being used, the employer is not free to leave helpersunprotected. The Secretary has established that Seibel was not using a\”proper\” degree of protection.III. Seibel’s Reliance upon Previous OSHA InspectionsSeibel’s brief raises a defense based on several earlier OSHAinspections, which did not result in any citations concerning eyeprotection for the helpers. President Seibel states in the brief:Since the other, much more experienced, compliance officers in morecomprehensive inspections did not find any safety problems whatsoever inthis area, and since the manner of assembling these battery cases iscontinuous throughout the work day and has not changed in forty years, Iwould assume [that Compliance Officer DeLeys’] inexperience and obviouslack of immediate knowledge led him to an erroneous conclusion.Also, at the hearing, President Seibel had testified:[I]n all the previous inspections[,] and we have had more thorough andcomprehensive inspections the last 2 or 3 times than what we receivedthis particular time, and none of those operations which have notchanged, have they ever minutely mentioned anything at all concerningour assembly and welding operations. Regarding shading for the fitters,in no way did they mention anything regarding eye protection for the twoman operation and they have viewed that operation completely.In short, Seibel infers from the uneventful prior inspections that theremust have been no hazard, and Seibel asks us to draw the same inferenceagainst OSHA, despite the cited OSHA standard’s presumption of a hazard.We should not, Seibel basically contends, hold the employer liable whereOSHA had earlier opportunities to issue citations but chose not to do so.In other cases, the Commission has rejected essentially the samecontention, that is, that the failure of OSHA to cite certain conditionsdeprived the employer of notice that those conditions were eitherhazardous or in violation of OSHA standards. The Commission hascautioned employers against freely drawing such inferences fromuneventful inspections, however, for \”[c]ertainly, an employer isrequired to comply with a standard regardless of whether it haspreviously been informed that a violation exists.\” Simplex Time RecorderCo., 12 BNA OSHC 1591, 1596, 1984-85 CCH OSHD ? 27,456, p. 35,572 (No.82-12, 1985). In Columbian Art Works, Inc., 10 BNA OSHC 1132, 1133, 1981CCH OSHD ? 25,737, p. 32,102 (No. 78-29, 1981), the Commission morepointedly warned: \”OSHA’s failure to issue a citation following aninspection does not grant an employer immunity from enforcement ofapplicable occupational safety and health standards.\”That the employer lacked notice of any noncompliance was the argumentrejected in Columbian Art Works and a related case, GAF Corp., 9 BNAOSHC 1451, 1457, 1981 CCH OSHD ? 25,281, pp. 31,246-47 (No. 77-1811,1981). These cases implicitly rule against deducing from uneventfulprior inspections that particular operations are nonhazardous. See alsoInternational Harvester Co. v. OSHRC, 628 F.2d 982, 985 n.3 (7th Cir.1980) (earlier failure to cite for violation of a particular standard isnot a decision that the employer was complying). Cf. Cedar ConstructionCo. v.OSHRC, 587 F.2d 1303, 1306 (D.C. Cir. 1978) (\”[w]e believe thatrecognizing such a right [to rely on uneventful prior inspections] woulddiscourage self-enforcement of the Act by businessmen who have fargreater knowledge about conditions at their workplaces than do OSHAinspectors\”).In Cardinal Industries, 14 BNA OSHC 1008, 1011, 1989 CCH OSHD ? 28,510,p. 37,801 (No. 82-427, 1989), the Commission upheld an employer’s viewthat \”the circumstances of the earlier OSHA inspections deprived it offair notice\” of a requirement to use safety belts. However, the earlierinspection had produced a citation for the very same condition as waslater cited, the earlier compliance officer had actually indicatedapproval of the very form of abatement that was later alleged to beinsufficient, and \”it was far from clear that any OSHA standardaddressed that particular fall hazard.\” 14 BNA OSHC at 1011-12, 1989 CCHOSHD at p. 37,801.[[9\/]]In Cardinal, there was a second lack-of-notice argument, which is moreclosely analogous to the situation now before us. The Commissionrejected this second argument because the earlier citation had notincluded the condition and had not prescribed its abatement. During theearlier inspection and during a follow-up inspection, the complianceofficer had been in the vicinity of the condition, but \”the mere fact ofMershon twice inspecting this area and not detecting violations does notexculpate Cardinal.\” 14 BNA OSHC at 1013, 1989 CCH OSHD p. 37,803. Inrejecting the employer’s argument, the Commission cited _Columbian ArtWorks_ and _Lukens Steel Co._, 10 BNA OSHC 1115, 1981 CCH OSHD ? 25,742(No. 76-1053, 1981)._Lukens_ is particularly pertinent to the case now before us:According to Respondent, this [prior] failure to issue a citation [for]the pouring platform supports an inference that the employees woreappropriate equipment. We disagree. The Commission’s function is todecide the cases before it, not to rule on the propriety of theSecretary’s actions in other cases . . . . Accordingly, we will notspeculate on the Secretary’s determinations in an earlier inspection.Further, because compliance with the Act is a continuing obligation, anemployer cannot deny the existence of or its knowledge of a cited hazardby relying on the Secretary’s earlier failure to cite the condition. . .. Consequently, the Secretary’s failure to cite conditions on thepouring platform during the earlier inspection does not \”exculpate\” theRespondent or preclude a finding that the violation now before us waswillful.10 BNA OSHC at 1126, 1981 CCH OSHD at p. 32,122 (case cites omitted).In essence, the mere fact of prior inspections does not give rise to aninference that OSHA made an earlier decision that there was no hazard,and does not preclude the Secretary from pursuing a later citation. Forthis reason, Seibel’s argument must be rejected. In this case, theSecretary could properly cite Seibel for violating the OSHA standard.Also, as we have discussed, the Secretary has established a violation.We turn therefore to Seibel’s other affirmative defenses, which thejudge did not consider.IV. _The Affirmative Defense of Greater Hazard _A. _Contentions of the Parties_Seibel contends that the wearing of shaded lenses would have endangeredthe fitters. \”[A]ny uncovered weld would have to be timed exactly… sothat the shaded lens[es] in the helmets are in position at the correcttime,\” otherwise the employer \”might have occur[re]nces of weld[ing]flash whereas now we have had none.\” Also, if a helper cannot \”see whathe is doing, the situation to him becomes more dangerous since thematerials are quite heavy with sharp corners and edges.\” [[10\/]]The Secretary contends that Seibel’s argument is \”mere speculation.\” Sheargues:Respondent seems to assert that having no shaded lenses at all would bebetter than having shaded lenses which possibly may not be in the properposition one hundred percent of the time… [T]he Commission hasrejected an all-or-nothing approach. Moreover . . . welders have oftenfailed to cover the weld arc with their hands even when helmets were notworn by the fitters. In view of the incidence of exposure presentlyoccurring, it cannot be said that respondent has demonstrated anincreased hazard.The Secretary further notes that, \”[b]ecause the welding operation is aregular part of respondent’s business, no excuse appears for Seibel’sfailure to seek a variance if it believed its claim had merit.\”B. _The Elements of the Defense_To establish a defense of greater hazard, an employer must prove that:(1) the hazards created by complying with the standard are greater thanthose of noncompliance, (2) other methods of protecting employees fromthe hazards are not available, and (3) a variance is not available orapplication for a variance is inappropriate. See Spancrete Northeast,Inc., 15 BNA OSHC 1020, 1022-23, 1991 CCH OSHD ? 29,313, pp.39,356-39,357 (No. 86-521, 1991). These three elements are nowwell-established in court precedent. Id. at n. 3 (listing cases). Anemployer’s proof of the unavailability or inappropriateness of avariance is particularly important. E.g., PBR, Inc. v. Secretary ofLabor, 643 F.2d 890, 895 (1st Cir. 1981). In this case, Seibel has notpresented any evidence of the unavailability or inappropriateness of avariance. On the contrary, the fact that the citation involves a regularand recurring operation at a permanent workplace strongly suggests thata variance application would be appropriate. Therefore, Seibel has notestablished this defense.V. _The Affirmative Defense of Infeasibility _A. _Contentions of the Parties_Seibel contends that shaded lenses at number 5 or greater would preventa fitter from seeing what he is doing. \”Our fitter has to measureaccurately,\” Seibel states. In testimony, President Seibel drew theconclusion that production errors could become more frequent if a fitterhad to put on shaded lenses because, as Seibel argues, a fitter has to\”handle the material with both hands to effect the required gaps,[overlapping] of [the] material, etc. as the tacks are performed.\” Thefitters function could not be supplanted by the use of jigs:[It is] out of the question except for the very smallest cases that areordered in quantities of twenty-five or more at a time. In that case, wedo utilize a jig fixture with one person to handle, assemble and tack acase.But, arguing the infeasibility of jigs for the two-person tack weldingoperations, Seibel asserts that \”[w]e have reviewed our battery caseassembly operation continuously for forty years,\” thereby implying thatSeibel’s own studies of its operations have continuously revealed thatthe functions now performed by helpers cannot instead be performed by jigs.The Secretary contends that Seibel, never having tried to use eitherappropriate eye protection or jigs, cannot simply assert that therewould always be operational impediments to the use of eye protection TheSecretary points out that:Even if . . . shaded lenses would have some effect on the fitter’sability to keep the [metal battery case] pieces lined up, there is noreal evidence as to what number of defective welds could be expected . .. . Nor is there evidence as to the consequences of misaligned pieces.Can the pieces be rewelded or must they be scrapped? Is there arejection\/repair rate that reasonably must be borne?Also: Since respondent’s expert agreed that even a 1.5 or 1.7 shade lensis ‘certainly better than a clear lens,’ and since there is no evidencenegating the possible use of at least a 1.5 or a 1.7 lens, a violationis established on this basis alone.Similarly, the Secretary claims that jigs probably could be used to a\”substantial\” extent:Seibel’s testimony indicates that the weight of the finished batterycases ranges from 30 pounds up to 1000 pounds, but there is no evidenceof the numbers of cases manufactured in various size categories. Thereis similarly no indication of the relative numbers of cases with anirregular configuration. These gaps in the evidence work againstrespondent, which bears the burden of proving its affirmative defense.The Secretary asserts that an employer must do at least as much as iscapable of being done, citing Bratton Furniture Manufacturing Co., 11BNA OSHC 1433, 1434, 1983-84 CCH OSHD ? 26,538, p. 33,858 (No. 81-799-S,1983).B. _The Case Law Defining Infeasibility_From the outset of the Act’s enforcement, the Commission has been facedwith employers’ claims that technological or economic problems precludedcompliance with cited standards. Until relatively recently, our casesruling on such claims stated that the employers must show that theabatement method required under the terms of the standard is\”impossible\” to use, in that it is physically impossible to implementor, if implemented, would preclude performance of necessary work. E.g.,M.J. Lee Construction Co., 7 BNA OSHC 1140, 1144, 1979 CCH OSHD ?23,330, p. 28,227 (No. 15094, 1979). In the mid-1980’s, however, theCommission re-examined its precedent, and the many court cases stemmingfrom it, and articulated a potentially less stringent test,\”infeasibility\”. Dun-Par Engineered Form Co. , 12 BNA OSHC 1949,1956-59, 1986-87 CCH OSHD ? 27,650, pp. 36,020-024 (No. 79-2553, 1986),rev’d in part, 843 F.2d 1135 (8th Cir. 1988) (\”Dun-Par I-A\”); compareSpancrete Northeast, Inc., 15 BNA OSHC at 1023, 1991 CCH OSHD at p.39,357 (employer might have basis to assert infeasibility in a casewhere impossibility had not been raised) . That is, an employer couldestablish a defense by showing that compliance would be \”infeasibleunder the circumstances.\” Cleveland Electric Illuminating Co., 13 BNAOSHC 2209, 2213, 1987-90 CCH OSHD ? 28,494, p. 37,761 (No. 84-593,1989), citing Dun-Par I-A, 12 BNA OSHC at 1953-56, 1986-87 CCH OSHD atpp. 36,021-024.The Secretary asserts in this case that Dun-Par I-A changed nothing butthe terminology.[[11\/]] In the Secretary’s view, infeasibilitynecessarily means impossibility since the Supreme Court, in AmericanTextile Manufacturers Institute v. Donovan, 452 U.S. 490 (1981)(\”ATMI\”), defined \”feasible\” as \”capable of being done.\”Thus the Secretary cautions in her brief herein:[I]f the Commission in Dun-Par intended the defense to be one ofimpracticality rather than impossibility, then this is a significantchange. See Century Steel Erectors v. Dole, [888 F.2d 1399 (D. C. Cir.1989)] (\”practical\” defined as \”relating to, or manifested in practiceor action; not theoretical or ideal\”; the term \”applies to things and topersons and implies proven success in meeting the demands made by actualliving or use\”; \”feasible\” means \”capable of being done or carried out\”).Contending that an impossibility test would be \”more consonant with theAct,\” the Secretary points to court decisions rejecting considerationsof \”difficulty,\” such as interference with existing work practices.One such decision is the law of the circuit of this case: Brennan v.OSHRC (Underhill Construction Corp.), 513 F.2d 1032 (2d Cir. 1975)(\”Underhill’). There, the cited standard prohibited materials storagewithin 10 feet of an open floor edge, but the cited subcontractormaintained that compliance would interfere with a customary workpractice, that of allowing reusable shoring and braces to overhang anopen floor edge so that cranes could lift them to other floors. TheSecond Circuit observed: \”[E]ven if the employer could raise [a] defense[of interference with work practices], on this record he did not meethis burden of showing that other means, e.g., strapping, chains,harnesses or the like, were not useable to permit ready crane removal ofthe shoring or steel braces.\” 513 F.2d at 1036 (emphasis added). Thecourt indicated that employers should carefully consider other \”useable\”work practices.We conclude that there is no conflict between the law of the circuit andCommission precedent concerning this element of the infeasibilitydefense; neither requires proof of impossibility. In fact, the view ofthe Second Circuit in Underhill is entirely consistent with theCommission’s in Dun-Par I-A. There, the question was \”whether Dun-Parestablished that guardrails could not have been installed or would havedisrupted the work to such a degree that there was no feasible way touse guardrails to protect Dun-Par’s employees.\” 12 BNA OSHC at 1959,1986-87 CCH OSHD at p. 36,027 (emphasis added). The Commission foundthat the use of guardrails was \”theoretically possible\” because it wasphysically possible to install them, but not \”feasible\” because theywere incapable of being used anywhere for a sufficient length of time toserve any practical purpose of protection. 12 BNA OSHC at 1959-60,1986-87 CCH OSHD at p. 36,027-028.[[12\/]] Compare Dun-Par EngineeredForm Co., 12 BNA OSHC 1962, 1966-67, 1985-87 CCH OSHD ? 27,651, pp.36,033-2 to 36,033-3 (No. 82-928, 1986) (\”Dun-Par II\”) (infeasibilitydefense rejected because the employer failed to demonstrate that thecosts would be \”unreasonable\” or that the use of guardrails would be\”unreasonable or senseless\”; also, the work practices involving use of acrane to lift shoring materials off a floor could have been altered toaccommodate guardrails).These cases show that employers must alter their customary workpractices to the extent that alterations are reasonably necessary toaccommodate the abatement measures specified by OSHA standards. That is,employers cannot generally avoid abatement by relying on industry customand practice alone. These cases do not stand for the proposition,however, that employers cannot rely on genuinely practical circumstancesrevealing the unreasonableness of an abatement measure. An abatementmeasure must be useable, during employees’ work activities, for itsintended purpose of protecting employees. If there is no way to use ameasure for its intended purpose without unreasonably disrupting thework activities, the mere fact that the measure’s installation isphysically possible does not in our view mean that we should compel theemployer to install the measure. This is why we do not see fit to applyin a literal manner the ATMI definition of feasibility–\”capable ofbeing done\”–and why we believe that the infeasibility test articulatedin Dun-Par I-A is the only reasonable test.C. _The Burden of Proof on the Feasibility of Alternative Measures _Early Commission precedent pertaining to affirmative defenses, inparticular, the affirmative defense of infeasibility and the affirmativedefense of greater hazard, which we have already discussed, establishedthat employers must either implement the abatement methods prescribed bythe OSHA standards or, to the extent the abatement methods proveunworkable because of infeasibility or greater hazard, implement analternative protective measure.[[13\/]] As is traditionally the case withaffirmative defenses, the burden of proof under the early Commissionprecedent rested with the cited employer who believed that both duties,that is, the duty to comply with the standard and the duty to providealternative protection, were incapable of performance because ofphysical conditions or work operations that were ongoing andinalterable, or counterproductive because of greater hazards. E.g. M.J.Lee, 7 BNA OSHC at 1144, 1979 CCH OSHD at p. 28,227. Cf., AstraPharmaceutical Products, Inc., 9 BNA OSHC 2126, 2131 n.16, 1981 CCH OSHD? 25,578, pp. 31,901 n.16 (No. 78-6247, 1981) (burden of proof inCommission adjudication’s is the preponderance of the evidence).In particular, to establish the affirmative defense that was known asthe impossibility defense, an employer had to show \”that (1) compliancewith the requirements of the cited standard either would be functionallyimpossible or would preclude performance of required work, and (2)alternative means of employee protection are unavailable.\” Id. Prior, tothe inspection in this case, however, the Commission’s decision inDun-Par I-A changed the defense from impossibility to infeasibility, aswe have discussed, and altered the burden of proof with respect to thetwo elements of the defense, in that they were divided between theparties. That is, if an employer showed the infeasibility of the citedstandard’s abatement measure, the burden of proof shifted to theSecretary to show the feasibility of an alternative abatement measure.Dun-Par I-A, 12 BNA OSHC at 1956-59, 1986-87 CCH OSHD at pp. 36,024-27.The question of whether the Commission should reconsider itsreallocation of the burdens of proof arose in April 1988 (prior to theNovember 1988 hearing in this case), when the United States Court ofAppeals for the Eighth Circuit, in response to the Secretary’s appeal ofDun-Par I-A, reversed the Commission’s ruling, holding that theCommission had erred in reallocating the burden of proof on alternativemeans of employee protection. Secretary v. Dun-Par Engineered Form Co.,843 F. 2d at 1137-40. The court’s decision does not apply directly tothe Seibel case, which arose in a different circuit, but in April 1989(after the hearing in this case, but before the parties filed theirarguments on review), a two-member Commission, in a decisionimplementing the court’s decision and remand order, confirmed onlytentatively that the commission would continue to follow itsreallocation of the burden of proof. Dun-Par Engineered Form Co, 13 BNAOSHC 2147, 2150, 1987-90 CCH OSHD ? 28,495, p.37,764 (No. 79-2553, 1989)(\”Dun-Par I-B\”).[[14\/]]The parties in this case were directed to file their review briefs inlate 1989. The Secretary relied on the Eighth Circuit’s opinion tojustify having not shouldered the burden of proof as to alternativemeasures for the purpose of the infeasibility defense:The Secretary disagrees with the Commission’s reallocation of the burdenof proof . . . . [T]he Commission overruled long-standing precedent that. . . the employer must show . . . there are no alternative abatementmethods. . . . For the reasons set forth in the Eighth Circuit’sdecision . . . the employer must retain the burden . . . .Seibel, appearing pro se and apparently unaware that the law hadchanged, argued only that it had established the infeasibility ofalternative measures.We now overrule Dun-Par I-A to the extent that it reallocates the burdenof proof regarding the infeasibility of any alternative measures. Forthe reasons given in the Eighth Circuit’s opinion, with which we agree,we hold that any employer seeking to be excused from implementing acited standard’s abatement measure on the basis of its infeasibility hasthe burden of establishing either that an alternative protective measurewas used or that there was no feasible alternative measure. As noted,the parties have tried and argued this time is though the Commissionnever reallocated the burden or prove. Therefore, there remains only thematter of making appropriate findings of fact and conclusion of law asto whether Seibel has met the burden of proof.D. _Remand for Factual Findings and Legal Conclusion on the AffirmativeDefense of Infeasibility._As our recitation of the facts at the outset of this opinion suggests,Seibel’s affirmative defense of infeasibility is highly fact-dependent;there are close questions of fact that are difficult to resolve. Theadministrative law judge did not examine the evidence and did not decidewhether the employer met the standard of proof regarding both elementsof the defense, i.e., the infeasibility of the abatement measurerequired by the cited standard (shaded lenses) and the infeasibility ofany alternative measure (jigs). Therefore, a remand isappropriate.[[15\/]] Compare, Kaspar Wire Works, Inc., 13 BNA OSHC 1261,1262, 1986-87 CCH OSHD ? 27,882 p. 36,554-55 (No. 85-1060, 1987) (remandfor the judge to address affirmative defenses apparently overlooked inhis decision).Seibel is appearing pro se and expresses a legitimate concern tominimize its litigation expenses. Therefore, we particularly note thatSeibel need not refile the arguments that have already been extensivelypresented to us and that our primary objective in remand is to put thecase back before an administrative law judge for the purpose ofresolution of Seibel’s infeasibility defense. Our intent is to assurethat Seibel and the Secretary are not denied the careful and completefactual review that is provided by orderly administrative procedures.VII. _Order_Accordingly, we reverse the decision of the judge and remand this casefor further proceedings consistent with this opinion, including anappropriate disposition of Seibel’s affirmative defense of infeasibility.Edwin G. Foulke,ChairmanVelma MontoyaCommissionerDonald G. WisemanCommissionerDated: August 9, 1991————————————————————————SECRETARY OF LABOR, Complainantv.SEIBEL MODERN MANUFACTURINGAND WELDING COMPANYRespondentOSHRC DOCKET NO. 88-0821_DECISION AND ORDER_This case arose under 29 USC, sec. 651 et seq. of the OccupationalSafety and Health Act of 1970 (the Act). As a result of an inspection bythe Occupational Safety and Health Administration [OSHA] of theRespondent’s premises, Citation #1 was issued charging that theRespondent violated sec. 5(a)(2) of the Act by the serious violation ofthe standard at 29 CFR 1910.252(e)(2)(i)(a), respectively item #3 ofCitation #1.On or about April 6, 1988, the Respondent filed Notice of Contest tosaid item 3 of Citation #1 and the penalty proposed therefor.The pertinent sections of the Act and the standard are appended to thisdecision under appropriate titles._PRELIMINARY_There is little disagreement about the facts. The alleged hazardconcerned employees who were engaged in welding. One employee, called a\”tacker\”, would do the actual welding on material held by anotheremployee, called a \”fitter\”. The tacker and fitter worked at a tablefacing each other with the part being assembled between them on thetable. The tacker wore a shield but the fitter wore no specialprotective glasses [some employees wore regular glasses]. The tackerwore a glove (on one hand). Both the tacker and fitter were about 24inches from the radiant energy of the welding, which was in plain viewof management personnel in the area._TESTIMONY_Seibel, the Respondent’s president, testified that in 3\/4ths of thewelds, the material itself is between the welding arc and the fitter;and that the gloved hand of the tacker is cupped over the weld entirelyin most cases …Tr. 57, 63-68. In his opinion, the operation was nothazardous and he is not aware of any cases of weld flash even though onefitter worked for 25 years and one for 8 years … Tr. 63.The Compliance Officer, DeLeys, testified that there were times when thefitter was not protected by either the glove or the material … Tr.100. He also conceded that his inspection was the first time he had everseen a 2-man welding operation…Tr. 28.The Respondent’s witness, Howard, testified that he had been a welderand fitter for 8 years and that the Respondent’s operation gives thefitter adequate protection… Tr. 77-82._RECORD OF INJURIES_The mere absence of any injury does not necessarily lead to theconclusion that there is no hazard; however, it is at least of someweight on the question of whether a hazardous condition exists. And thefewer the number of injuries and the greater the length of time withoutinjury are both of some consequence in answering that question. In theinstant case, there is uncontradicted testimony that the Respondent’smethod of welding had been carried on for some 25 years without anyaccident or injury to any employee. That record certainly supports theRespondent’s contention that its method of welding is nonhazardous._DISCUSSION _I was very favorably impressed by the Compliance officer (DeLeys).However, he had the burden of proof. He had to overcome the opinion ofthe Respondent’s president [Seibel] and the Respondent’s expert [Howard]who had both had years of experience whereon the Compliance Officer hadnever before seen a 2-man operation. In addition, the opinion of theRespondent’s witnesses was corroborated by the Respondent’s years ofoperation without injury or accident. It was simply too much to beovercome by the Compliance Officer.I find that the Complainant has not sustained the burden of proving thatthe Respondent violated the standard at 29 C.F.R. 1910.252(e)(2)(i)(a)._CONCLUSIONS OF LAW_1. At all times concerned, the Respondent was an employer engaged in abusiness affecting commerce within the meaning of the Act; and theOccupational Safety & Health Review Commission has jurisdiction over thesubject matter and the parties.2. The Complainant has not sustained the burden of proving theRespondent violated Sec. 5(a)(2) of the Act (See. 654)._ORDER_The whole record having been considered, and due consideration havingbeen given to 29 U.S.C. Sec. 666(j), it is ordered:1. Item 3 of Citation #1 is vacated, together with the proposed penalty.So ordered.FOSTER FURCOLOJudge, OSHRCDated: March 28, 1989Boston, Massachusetts_APPENDIX__THE ACT_1. Section 654 [section 5(a)(2)] Employer\” … shall comply withoccupational safety and health standards … \”2. Section 666 [section 17(b)] \”… employer who has received a citationfor a serious violation… of this Act … shall be assessed a civilpenalty of up to $1,000 for each such violation.\”3. Section 666 [section 17(k)] \”… a serious violation shall be deemedto exist … if there is a substantial probability that death or seriousphysical harm could result… unless the employer did not, and could not… know of the presence of the violation.\”_THE STANDARD_29 CFR 1910.252(e)(2)(i)(a): \”Helmets or hand shields shall be usedduring all arc welding … Helpers or attendants shall be provided withproper eye protection.\”————————————————————————FOOTNOTES:[[1\/]] The cited standard is 29 C.F.R. ? 1910.252(e)(2)(i)(a), whichprovides:Helmets or hand shields shall be used during all arc welding or arccutting operations, excluding submerged arc welding. Helpers orattendants shall be provided with proper eye protection.[[2\/]] We have amended the caption of this case to correct an apparenterror as to the employer’s name, which was originally docketed as\”Seibel Modern Manufacturing & Welding Company.\” The employer’s briefsand other submissions state the name as \”Seibel Modern Manufacturing &Welding Corp.\”[[3\/]] In testimony, the welders were called \”tackers,\” and the helpers,\”fitters.\” In this decision, however, we will use the terminology of thecited standard.[[4\/]] Lens shades are numbered (1.5 through 14) to designateprogressively deeper shades. Compliance Officer DeLeys and OSHA’stechnical support group, which he consulted, believed that Seibel’shelpers should wear shade number 10, which Seibel’s welders wore. ExpertHoward testified that shade number 10 is appropriate for welders who areusing 85-90 amp welding equipment, which is the kind of weldingequipment that Seibel was using. Expert Howard and President Seibel bothbelieved, however, that even shade number 5 would be too dark forSeibel’s helpers to wear.[[5\/]] President Seibel further testified that he was \”not sure there isa magnified ruler that can be used in this particular instance.\” Theiruse was not further explored.[[6\/]] The guide to which the latter provision refers recommends a shadenumber 12 for \”[g]as-shielded arc welding (ferrous)\” using electrodesranging from 5\/32 down to 1\/16 inch. Seibel’s gas-metal arc welding orMIG (metal inert gas) welding process employed a smaller electrode thanthe guide mentions. Seibel’s electrode was .035 inch, closer to 1\/32than 1\/16 inch (which is .0625 inch). Compliance Officer DeLeys and theOSHA technical support group that he consulted therefore recommended ashade number 10, rather than shade number 12, as specified by the guide.[[8\/]] Moreover, aside from the guidance given in the standard, in thiscase there is opinion evidence suggesting that this employer couldactually have recognized the need for shaded lenses. Seibel’s expertwitness testified that shade numbers 1.5 or 1.7 would be advisable forthe helpers. (Compliance officer DeLeys testified, however, that hewould not have accepted such a slight shade.) Also, president Seibel hadalready realized that the welders needed shaded lenses, and UnionCarbide, the manufacturer of Seibel’s welding equipment, recommendedshaded lenses for \”observers\” of welding.[[9\/]] See also Hamilton Die Cast, Inc., 11 BNA OSHC 2169, 2172, 1984-85CCH OSHD ? 26,983, p. 34,690 (No. 79-1686, 1984) (”[i]n view of thegeneral nature of the cited standard and the lack of any expresslanguage specifically addressing die casting machines, HDC cannotreasonably be said to have been on notice of a requirement toguard….once the Secretary had informed HDC that guarding was notneeded at such times\”).[[10\/]] President Seibel first made this point in argument, but it findssome support in his testimony and other record evidence. He haddiscussed the matter of a helper’s not being able to see through shadedlenses, and he had described the unwieldiness of the metal battery casepieces their large size, considerable weight, awkward shape and bowedcondition. He had also offered into evidence photographs of fittersholding large, flat and thin pieces of metal with sharp corners andragged edges. The photographs show that the fitters can sometimes weargloves but do not always do so: for example, one photograph shows afitter wearing one glove (on the hand holding a metal piece) and havingone hand bare (holding pliers onto another metal piece); anotherphotograph shows the same fitter holding a metal piece with his bare hands.[[11\/]] The Secretary did not appeal the part of Dun-Par I-A that dealtwith an employer’s burden to establish the infeasibility of theabatement method specified in a cited standard.[[12\/]] Therefore, the Commission decided that it would be unreasonableto limit the employers’ defense to impossibility: \”Strict application ofan ‘impossibility’ defense does not accommodate considerations ofreasonableness or common sense, or reflect the strong sense of thepractical.\” 12 BNA OSHC at 1955, 1986-87 CCH OSHD at p. 36,023.[[13\/]] See Universal Sheet Metal, 2 BNA OSHC 1061, 1062, 1973-74 CCHOSHD ? 18,163, p. 22,341 (No. 657, 1974) (safety belts used inalternative to guardrails); George A. Hormel and Co., 2 BNA OSHC 1190,1192-93, 1974-75 CCH OSHD ? 18,685, pp. 22,582-83 (No. 1410, 1974)(safety harness as alternative to guardrails); American Bridge, 2 BNAOSHC 1222, 1223, 1974-75 CCH OSHD ? 18,702, p. 22,593 (No. 2249, 1974)(factual indication that neither cited abatement nor alternative wasusable); Cimpl Packing, 2 BNA OSHC 1436, 1437 n.2, 1974-75 CCH OSHD ?19,127, p. 22,847 n.2 (No. 1987, 1974) (failure to explore use ofalternative measure); Garrison & Associates, Inc., 3 BNA OSHC 1110,1974-75 CCH OSHD ? 19,550, p. 23,344 (No. 4235, 1975) (cited standard’sadditional measure– hand tools–used in alternative to required machineguarding, which might have been unusable); Central Steel and Tank, 3 BNAOSHC 1711, 1712, 1975-76 CCH OSHD ? 20,172, p. 24,017 (No. 2346, 1975)(\”[e]ven accepting the validity of this claim\”–that the citedstandard’s methods of machine guarding could not be used–\”the recorddoes not establish that no practical means of protecting against thehazard exists\”). See also Taylor Building Associates, 5 BNA OSHC 1083,1085 n.6, 1977-78 CCH OSHD ? 21,592, p. 25,910 n.6 (No. 3735, 1977) andother cases cited therein, especially Grossman Steel & Aluminum Corp., 4BNA OSHC 1185, 1975-76 CCH OSHD ? 20,691 (No. 12775, 1976) , andAnning-Johnson Co., 4 BNA OSHC 1193, 1975-76 CCH OSHD ? 20,690 (No.3694, 1976), which involve the multiple-employer worksite defense.[[14\/]] Former Commissioner Arey stated: \”I leave for another day thequestion of whether current Commission precedent is correct; or whetherthe Eighth Circuit’s position should be adopted. Commission precedentstill places the burden on the Secretary of showing a feasiblealternative means of protection when literal compliance with a standardis infeasible.\” Former Chairman Buckley stated: \”I concur withCommissioner Arey’s disposition of this case, given the mandate imposedon the Commission by the Eighth Circuit . . . . \” and \”[I]t is withreluctance that I follow the Eighth Circuit’s decision . . . . \” 13 BNAOSHC at 2151 & 2153, 1987-90 CCH OSHD at pp. 37,766 & 37,767.[[15\/]] \”[T]he Commission is the fact-finder, and the judge is an arm ofthe Commission for that purpose.\” Accu-Namics, Inc. v. OSHRC, 515 F.2d828, 834 (5th Cir. 1975), cert. denied, 425 U.S. 903 (1976) (emphasis inthe original). The Commission has authority to make factual findingswhere the administrative law judge did not, Accu-Namics, 515 F.2d at834, but ordinarily an administrative law judge does resolve thesefactual issues first and the Commission then exercises a reviewfunction. This bi-level procedure within an administrative agency wasconceived to benefit the parties, and the benefits of it areparticularly evident in cases of close questions of fact. Cf., Ed TaylorConstr. Co. v. OSHRC, 931 F.2d 1458, 1462-64 (11th Cir. 1991) (citationitem vacated because vacancies on the Commission denied the employer anopportunity for Commission review of the administrative law judge’sfactual findings, under the preponderance of the evidence test). At thenext level of adjudication–review by a federal court of appeals–thefactual findings of the Commission are conclusive as long as they aresupported by substantial evidence. 29 U.S.C. ? 660(a).”