Docket No. 77-1796

UNITED STATES STEEL CORPORATION

OSHRC Docket No. 77-1796

Occupational Safety and Health Review Commission

June 30, 1982

[*1]

Before ROWLAND, Chairman; CLEARY and COTTINE, Commissioners.

COUNSEL:

Office of the Solicitor, USDOL

Marshall H. Harris, Reg. Sol., USDOL

J. T. Carney, General Attorney, United States Steel Corp., for the employer

OPINION:

DECISION

BY THE COMMISSION:

A decision of Administrative Law Judge William E. Brennan is before the Commission pursuant to section 12(j) n1 of the Occupational Safety and Health Act of 1970, 29 U.S.C. § § 654-678 ("the Act"). United States Steel Corporation ("USS" or "Respondent") takes exception to the judge's affirmance of a citation issued in connection with its open hearth furnance operation in Fairless Hills, Pennsylvania. The citation and complaint issued by the Secretary of Labor ("the Secretary") allege that USS failed to comply with section 5(a)(1) of the Act, the "general duty clause," n2 in that:

A safe and healthful workplace free of recognized hazards likely to cause death or serious physical harm is not provided to employees in the open hearth in that employees are required to change oxygen lances atop the furnaces with molten metal in the furnaces. This process involves hazards associated with water becoming entrapped below molten metal.

- - - - - - - [*2] - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n1 29 U.S.C. § 661(i).

n2 Section 5(a)(1), 29 U.S.C. § 654(a)(1), provides:

Sec. 5(a) Each employer --

(1) shall furnish to each of his employees employment and a place of employment which are free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees.

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

The issues on review are whether the administrative law judge erred in deciding the case on a theory other than that agreed upon at the outset of the hearing, and whether a preponderance of the evidence demonstrates a violation of the general duty clause. n3

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n3 Former Commissioner Barnako's direction for review requested submissions from the parties on two issues: (1) whether the trial judge committed reversible error in affirming that item of the citation which alleged that employees required to change oxygen lances are exposed to hazards associated with water becoming trapped below molten metal, and (2) if the judge did not err, whether he assessed a proper penalty. In its petition for review USS argued, in addition, that the judge erred in affirming a second item of the citation alleging that employees involved in lance changes could be knocked off balance and onto the roof. That item was not directed for review and is therefore not before us. See Henry C. Beck Co., 80 OSAHRC 67/D13, 8 BNA OSHC 1734 n.2, 1980 CCH OSHD P24,590 at p. 30,169-3 n.2 (No. 77-963, 1980).

[*3]

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

The investigation which led to the citation in this case was conducted in response to an employee complaint to the Occupational Safety and Health Administration ("OSHA") of unsafe work conditions. Specifically, millwrights at the Fairless Works objected to a newly instituted work rule which required that oxygen lances be replaced in furnaces containing molten metal. Previously the practice had been to change inoperative lances before molten metal was added to a furnace. The millwrights told the OSHA compliance officer that the new procedure was dangerous because it could cause water to become entrapped in molten metal and explode. n4

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n4 The change in work practices was the subject of an arbitration proceeding. The board of arbitration declined to rule on the merits of the millwrights' complaint, but determined that USS had failed to comply with its collective bargaining agreement by attempting to negotiate the terms of the new work rule directly with the millwrights rather than through the company's Joint Health and Safety Committee.

[*4]

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

Because the issues on review turn largely on questions of evidence, and that evidence tends to be highly technical in nature, we begin our discussion with a description of the open hearth furnace operation, according to the evidence developed below. USS's nine open hearth furnaces resemble large bathtubs, each measuring 60 to 90 feet by 12 to 20 feet, with an interior depth of 15 to 16 feet. They are constructed of heat-resistant basic bricks mounted on steel rods and are protected from the intense heat by a network of pipes which carry cooling water over the structures, particularly around doors and other apertures. Furnace doors are fitted with "wicket holes" through which the millwrights can observe the contents to the furnace and through which samples of molten metal are extracted for analysis.

Steel is produced in the open hearth operation at a rate of about 750 "heats" per month. A heat begins with the charging of limestone, scrap metal, and ore into the empty furnace. Molten iron is then poured in. The initial interaction of hot iron, ore, and scrap metal causes a violent boiling action [*5] which can send molten metal flying out of the furnace. For this reason "Danger" signs are posted on the charging floor in front of the furnace before hot iron is added, and employees are kept at least 40 feet from the furnace doors until the foreman determines that the bath has stabilized. This violent period usually lasts one to two hours. Thereafter the mixture "cooks" anywhere from four to eight hours, until the desired grade of steel has been produced and is tapped off. n5 As the mixture cooks, impurities called slag rise to the surface. The furnaces are designed so that some of the slag is removed through openings in the furnace, but a layer three to four inches thick generally remains throughout the heat. Temperatures inside the furnace during a heat range from 2300 to 2900 degrees Farenheit.

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n5 The administrative law judge found as fact that the bath is unstable from the moment molten iron is added to the ore and scrap until refined steel is tapped off and that this entire period lasts about one and one-half hours. This is error. The record establishes that the period of instability or turbulence to which the judge referred is followed by at least three to four hours of "cooking" time during which the bath remains stable unless ore is added or substantial water leaks develop.

[*6]

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

In the mid-1950s it was discovered that the heat time could be reduced by about half if oxygen is blown into the bath as it cooks. The oxygen is pumped into the furnace through "lances," or "jets," which are 18-foot steel pipes, n6 six and one half-inches in diameter, weighing approximately 900 pounds. Each furnace is equipped with two or three lances, suspended vertically by overhead crane through eleven-inch holes in the roof. Oxygen enters the lance at its top end, which protrudes above roof level, and is blown out through oxygen ports in the bottom, the "head." Each lance is protected from the intense heat by cooling water which is circulated through the pipe's interior, entering and exiting through two two-inch ports at the top of the lance. A fully functional lance holds 18 gallons of water.

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n6 The judge stated in his decision that the lances are 16 feet long. However, Exhibit C-17, which is a blueprint of one of USS's furnaces, shows that the overall length of a lance is 18 feet.

- - - - - - - - - - - [*7] - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

The lances are subject to two types of malfuction. The oxygen ports can become clogged, and substantial water leacks can develop. In either event, the non-operational lance is pulled up and away from the metal bath, its oxygen and water supplies are turned off at the source, and the flow of oxygen to the remaining lance or lances is increased. Until 1975, if all lances in a furnace failed during the heat, the heat was simply continued without benefit of added oxygen at the slower rate necessary to that process. In May 1975, however, USS directed the millwrights to change one lance if all lances in a furnace failed, even while the furnace contained molten metal, so long as the foreman had determined that the bath was stable.

The procedure for changing a lance is essentially the same whether or not the furnace contains molten metal. The replacement lance is suspended by a cable and moved by a crane from the charging floor to the area above the furnace where three men standing on girders and beams directly above the roof push and pull the suspended pipe into the lance hole. Once the tip of the lance has been meneuvered [*8] into position, its water and oxygen supplies are turned on. The lance is then lowered until its head is directly above but not touching the bath. A man standing on the charging floor watches the lance's descent through the wicket hole and signals to the crane operator when the lance has reached its proper depth. Due to the great weight of the lance and the limited clearance above the furnace roof, it is not uncommon for the lance to bump into and break water pipes on the roof, causing water to leak into the furnace.

As the bath cooks it is in a constant state of upheaval, much like a pot of boiling candy. Stringers of molten metal are sometimes tossed up from the surface and even out of the furnace itself. This boiling action can be significantly increased by water leaks, which are an unavoidable accompaniment to the process. Water pipes are often broken by unmelted scrap and ore as they are tossed about in the furnace, or by the oxygen lances in transit over the roof. In addition, the lances themselves can develop water leaks. When large amounts of water leak into a furnace the normal agitation of the bath is increased. This is particularly true when the water is released [*9] beneath the surface of the bath, as in the case of a lance head becoming submerged in molten metal and the casing being melted away before the water supply is turned off. In addition to the increased agitation caused by water leaks, the steam which results pours out of the furnace, pushing with it flames and smoke.

Another cause of increased turbulence in the bath is the addition of limestone or ore to the molten metal at an advanced stage of the heat. When ore is added to reduce the temperature of the bath or to adjust the grade of steel being produced, the initial interaction between the ore and the molten metal releases gases which cause an intense boiling action capable of tossing hot metal out of the furnace.

The record establishes that there are three reasons why a lance change would not be made during the period of instability caused by addition of ore to molten metal. First, instability in the bath is readily apparent to the melter foreman and first helper, whose job it is to assess the stability of the bath at all stages of the heat. Under the new procedure a molten metal lance change would be prohibited over an unstable bath. Second, lance failure is normally manifested [*10] by reduced temperature in the bath. The bath cannot be tested for temperature if it is unstable. Therefore, lance failure and the need to replace a lance will not be apparent until the bath has stabilized. Third, a lance change is only made if all the lances are inoperative. In such a situation, however, the temperature of the bath drops and it would not be necessary to add ore to reduce the temperature.

On the first day of the hearing the parties and the administrative law judge agreed that the issue to be tried was whether the practice of changing oxygen lances over molten metal will result in entrapment of water and explosion. USS conceded that if water is trapped below molten metal an explosion can result, but denied that the practice of changing lances over molten metal will cause entrapment of water.

During the five days of the hearing frequent reference was made to the limited scope of the Secretary's charge:

The Judge: I think the pleadings and the documents within the docket file pretty much reflect at least part of that position. That is, that the Respondent does not agree that there is a danger of water entering intol the molten metal because of the practice [*11] of changing the oxygen lances; but perhaps we have saved some time by stipulating that getting water into the metal is a recognized hazard.

The Secretary: Well, once water is into the metal and entrapped in the metal, that is a recognized hazard. Really, the issue rendered by that item is whether or not the practice of changing lances will result in water becoming entrapped.

The Judge: That is the way I read the file. All right. Any further stipulations?

* * *

The Secretary: It is the theory of the Government that the recognized hazard which is an explosion, can be the result of changing oxygen lances during the time molten metals are in the furnace.

* * *

The hazard, of course, is the potential of an explosion. Cited in the citation is the practice which by its nature may lead to the possibility.

* * *

The Judge: Millwrights don't want to be up on that roof if there is a chance of a tube of water falling into that hot metal because they are convinced . . . that you can wind up with an explosion.

* * *

[T]he question there is when the lance is being changed with hot metal in the furnace, is there a realistic possibility of an explosion or other type of occurrence because [*12] of that process which would endanger the Millwrights that are doing the job.

* * *

There has to be encapsulation for explosion. We then come to the question of what are the realistic possibilities of such encapsulation because of the oxygen lance procedure?

* * *

What I am asking you . . . to do is try and prove to these Millwrights that their fear of a lance exploding is not founded on fact.

On occasion USS objected to evidence on the ground that it was not probative of an issue in the case:

The Union: Now, can you relate some of the causes of a violent reaction insofar as they are related, let's say to a water problem?

USS: Your Honor, I object. I think again we are going beyond the scope of direct examination. I also object on grounds of relevancy because that is not the issue in this case.

The Court: Technically you are right, Mr. Carney.

* * *

The Union: Can you describe for us the type of damage caused by the addition of water or after the addition of hot metal and such violent reactions?

USS: Your Honor, I object again. I think we are just going far afield in terms of talking about -- in extensive detail a reaction which is not at issue in this case.

In deciding [*13] the case, however, the judge did not rule on the issue framed by the parties. He acknowledged that the question for decision was whether the changing of lances over molten metal would result in entrapment of water, causing explosions, and that "Complainant's evidence, in major part was aimed at establishing this possibility whereas Respondent's evidence was directed to showing no such realistic possibility exists." Nevertheless, he concluded that the record when taken as a whole shows that the real danger posed by the lance change procedure is that it brings employees within range of molten metal eruptions caused by factors other than the lance change itself. "The chance of explosion or 'violent eruptions,' from many unpredictable water sources, with the risk of partial roof damage or collapse, and the spraying of molten steel, is real, based on the total evidence of the record."

On review USS argues that it was denied due process of law when the judge changed the issue in the case after the hearing and without affording USS an opportunity to present argument or evidence on the new issue. USS points to the passages in the record quoted above in which the judge stated that the [*14] hazard at issue was explosion caused by the lance change procedure and notes that although the judge states that the broader issue was specifically tried by the parties, he cites no evidence in the record that any issue other than that agreed upon at the outset was being litigated. Finally, USS points out that its brief to the judge was limited to the question of whether the lance change procedure could cause entrapment of water.

We agree with USS that the judge acted improperly in finding a violation based on possible eruptions of molten metal from causes other than a lance change. The citation was limited to a potential hazard resulting from a lance change, and statements made at the hearing by all parties and the judge establish that the parties did not intend to expand the charge beyond that contained in the citation. It was therefore error for the judge to decide the case on a different issue. "It is patently unfair for an agency to decide a case on a legal theory or set of facts which was not presented at the hearing." National Realty and Construction Co. v. OSHRC, 489 F.2d 1257, 1267 n.40 (D.C. Cir. 1973).

We recognize that the millwrights are genuinely concerned [*15] for their safety and that their concern extends beyond the issue tried by the parties. n7 However, the Commission's decision must be based on evidence of record as it pertains to the Secretary's theory of the case. The Secretary chose to limit the charge to the hazard of entrapment of water in molten metal as caused by the lance change procedure. During the five days of the hearing the judge and the parties repeatedly referred to the limited nature of the charge. Evidence of nonentrapment hazards created by the ordinary operation of the furnace was relevant to the agreed-upon charge because it provided background information necessary for an understanding of this highly techncical operation, and because the millwrights who provided much of this testimony did so in response to the questions regarding the basis of their fears of the molten metal lance change procedure. Under these circumstances USS did not receive fair warning that a broader issue was being tried.

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n7 Although we do not rule on the broader issue framed by the judge, we note several errors in the judge's analysis of the evidence, which cast doubt on the accuracy of his conclusion that millwrights performing a molten metal lance change would be exposed to hazards arising out of the ordinary operation of the furnace. The record establishes that there is a period of extreme turbulence immediately after the addition of molten iron to the scrap and ore, that this period lasts approximately one and one-half hours, and that thereafter the mixture cooks for several more hours in a relatively stable manner unless certain specified events occur. The judge, however, mistakenly concluded that the one and one-half hours of extreme turbulence constituted the entire cook time, so that millwrights going up on the roof after addition of molten iron are necessarily exposed to a highly agitated bath. See note 5 supra. He also concluded that until the institution of the new lance change procedure millwrights were never required to go up on the roof after hot iron had been added to the furnace. However, as indicated at note 9, infra, there is unrebutted testimony in the record that millwrights have traditionally gone up on the roof, without objection, to repair door frames while the furnace contains molten metal. The judge also disregarded entirely the testimony of the two expect witnesses on the ground that it related "chiefly to the scientific physics of explosions as they might be caused by an oxygen lance falling into a furnace after the addition of hot metal. . . ." However, inasmuch as the experts' testimony was relevant to the case, the judge's reason for disregarding it is not valid. The judge also improperly rejected in large part the testimony of USS's Division Superintendent regarding precautionary measures that would be taken before and during a lance change on the ground that this witness was not required to change lances over molten metal himself.

[*16]

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

III

Section 5(a)(1) requires employers to provide their employees workplaces "free from recognized hazards that are causing or are likely to cause death or serious physical harm." Numerous Commission and court decisions have interpreted this language and have given various formulations of the elements the Secretary must prove in order to establish that an employer violated section 5(a)(1). See, e.g., Baroid Division of NL Industries, Inc. v. OSHRC, 660 F.2d 493 (10th Cir. 1981); Whirlpool Corp., 79 OSAHRC 32/A14, 7 BNA OSHC 1356, 1979 CCH OSHD P23,552 (No. 9224, 1979), vacated and remanded on other grounds, 645 F.2d 1096 (D.C. Cir. 1981). In Bomac Drilling, Division of TRG Drilling Corp., 81 OSAHRC 45/A2, 9 BNA OSHC 1681, 1981 CCH OSHD P25,363 (No. 76-450, 1981), the Commission expressed the elements of a section 5(a)(1) violation in a manner that differed from most prior court and Commission decisions by including the reasonable foreseeability of an incident as a distinct element of a violation:

In order to establish a section 5(a)(1) violation, the Secretary must show [*17] that an employer failed to render its workplace free from a hazard that is recognized, the occurrence of an incident was reasonably foreseeable, and the likely consequence in the event of an incident was death or serious physical harm to its employees . . . . The Secretary must also demonstrate that there were feasible means available to abate the hazard.

9 BNA OSHC at 1691, 1981 CCH OSHD at p. 31,547 (citations omitted). In a concurring opinion in Bomac, Commissioner Cleary criticized the inclusion of the reasonable foreseeability of an incident as a separate element of a section 5(a)(1) violation. He noted that the facts relied on by the majority to conclude that an incident was reasonably foreseeable were the same as those on which they found that the employer had failed to free its workplace from a recognized hazard. Thus, he considered that including the reasonable foreseeability of an incident as a separate element of a violation simply added an unnecessary complication to the analysis of section 5(a)(1) cases.

We conclude that the reasonable foreseeability of an incident should no longer be considered a separate element of a secion 5(a)(1) violation. In addition [*18] to agreeing with Commissioner Cleary's rationale in Bomac, we note that the United States Court of Appeals for the Second Circuit has specifically disapproved of the Commission's application of "reasonable foreseeability" to a section 5(a)(1) case. Pratt & Whitney Aircraft v. Secretary of Labor, 649 F.2d 96 (2d Cir. 1981). Moreover, other courts of appeals that have decided section 5(a)(1) cases have done so without resorting to a discussion of whether an incident was reasonably foreseeable. Baroid Division of NL Industries, Inc. v. OSHRC, supra; St. Joe Minerals Corp. v. OSHRC, 647 F.2d 840 (9th Cir. 1981); Whirlpool Corp. v. OSHRC, 645 F.2d 1096 (D.C. Cir. 1981); Empire Detroit Steel Division, Detroit Steel Corp. v. OSHRC, 579 F.2d 378 (6th Cir. 1978); Getty Oil Co. v. OSHRC, 530 F.2d 1143 (5th Cir. 1976); National Realty & Construction Co. v. OSHRC, supra. The disparity between the formulation stated in Bomac and that stated by such a large number of courts of appeals is a potential source of confusion and fruitless litigation. Accordingly, we overrule Bomac to the extent it held that the reasonable foreseeability of an incident [*19] is a distinct element of a section 5(a)(1) violation.

The citation alleges that USS failed to render its workplace free of the hazard of explosions resulting from water becoming entrapped below molten metal as a result of a lance change. USS contends that a lance change over molten metal cannot lead to water entrapment and a resultant explosion. Although the administrative law judge did not rule on this issue, we need not remand for further findings of fact. The Commission, not the administrative law judge, is the ultimate trier of fact. Accu-Namics, Inc. v. OSHRC, 515 F.2d 828 (5th Cir. 1975), cert. denied, 425 U.S. 903 (1976).

Both USS and the Secretary produced expert testimony on the question of entrapment explosions, Geoffrey Belton for the Secretary, and Chester Bieniosek for USS. Although they disagreed sharply in several respects, Belton and Bieniosek were in complete agreement that water released below the surface of the bath will normally vaporize and come to the top rather than become entrapped. An explosion due to entrapment of water occurs only if water is subjected to extreme heat while caught behind a solid barrier which prevents its escape as [*20] it vaporizes and expands. USS's expert, Bieniosek, explained how an entrapment explosion of this type can occur in an open hearth furnace. Water, particularly ice or snow, is sometimes mixed in with the ore charged at the bottom of a furnace. Whem molten iron is poured onto the ore and scrap, a "skull" of solid metal can form temporarily between the water at the bottom of the furnace and the hot iron. As the heat transfers through the skull, the water beneath evaporates, expands, and builds up sufficient pressure to blow through the skull. Also, water is sometimes trapped inside the batteries mixed in with the scrap metal. Bieniosek testified that the potential for this kind of explosion is one of the reasons the danger sign is posted on the charging floor until the melter foreman is certain that the mixture in the furnace has stabilized.

The two experts also agreed that if a lance containing water were to fall into the bath, if its water ports were to become blocked, and if water were trapped inside, the lance could explode. Further, both experts stated that it is not possible to calculate with mathematical certainty the rate and force with which steam would issue from [*21] the lance or the rate at which slag would solidify in the ports. They reached differing conclusions, however, as to whether water could actually become entrapped in a fallen lance.

Belton, the Secretary's expert, hypothesized that a lance could fall into the furnace and water could become trapped inside if the lance fell through the lance hole at about a 30-degree angle n8 or if it fell through a large hole in the roof caused by a collapse of the brick and metal structure. The lance would descent into the bath like a falling spear, bouy back up, and come to rest floating horizontally in the slag layer. The slag would begin to solidify in the ports immediately upon contact with the lance because the lance would be much colder than the slag. In Belton's opinion this solidifying process would occur more rapidly than the water inside the lance would turn to steam. Consequently the water would be trapped inside as it vaporized and expanded, and eventually the lance would explode.

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n8 The lance must fall at an angle in order to clear the roof, because the lance is approximately 18 feet long and the furnace interior is only 15-16 feet deep. On the other hand, the lance must not fall at too oblique an angle or it will be caught in the lance hole, which is only 11 inches in diameter.

[*22]

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

In Bieniosek's opinion, the chance of this happening is "practically zero" because steam would rush out of the ports with sufficient force and speed to prevent slag from blocking the ports. Bieniosek stated categorically that the extremely high temperature and large quantity of molten metal surrounding the lance would cause the water inside to vaporize so quickly and with such force that it would blow out of the ports before they could become blocked. Even if it were possible for a layer of slag to obstruct the ports completely, that slag would be so weak and thin that the steam could easily break through. He stated that one reason for the slag's inability to form a strong barrier in the ports is that it is a complex material that does not solidify at a given temperature, but "goes through a range of temperatures during the solidification process which means for a considerable time there is a mixture of solids and liquid in the slag."

We find Bieniosek's testimony more convincing than Belton's for the following reasons. Belton predicated his hypothesis on several factors not present at the cited [*23] furnaces. He assumed a slag depth of eight of twelve inches, but it was the uncontradicted testimony of USS's plant superintendent that the maximum slag depth is three to four inches. He assumed the water ports are one and one-half inches in diameter, but blueprints in evidence show that the ports are two inches in diameter. His hypothesis required that the lance fall through the lance hole at a 30-degree angle or through a large hole in the roof, but he specifically stated he had made no effort to calculate the likelihood of a lance falling at the requisite angle, and it was the uncontradicted testimony of the Fairless Division Superintendent and the Open Hearth General Foreman of Maintenance that a lance could not fall into the furnace due to a collapse of the roof. Both witnesses testified that the network of steel rods on which the bricks are mounted would make it impossible for large sections of roof to collapse, and that a lance swinging above the roof would be caught by the pipes and genders above the roof. Moreover, Belton did not state his opinion with any degree of certainty. Although he stated on direct that it is "very likely" that the water ports would become clogged [*24] with solidified slag before steam could escape, on cross examination he conceded that "in general [I] think the solidification is probably less likely than the evolution of steam from the nozzles." Further, Bieniosek had more professional experience in the matters at issue than Belton. Belton stated that he had no direct experience with oxygen lances, had never personally observed a lance change, and had no professional experience working with the interaction of liquid water and steel. Bieniosek, on the other hand, described extensive experience in the interaction of oxygen and water with molten metal and had observed oxygen lances in operation for 20 years. He stated that during that time he had never seen or heard of an explosion anywhere caused by a fallen lance. Finally, he gave his expert opinion without reservation.

The two experts also testified regarding the possibility that large quantities of water floating on top of the bath could cause an entrapment explosion. According to Belton, if water were to pour into the furnace at a rate of 250 gallons per minute, either from a broken lance or from pipes broken by the lance in its progress over the roof, not all the water [*25] would vaporize. Instead, a layer of steam would form on top of the slag, and liquid water would float on that. If a lance, or bricks or pipes knocked loose by a lance, were then to fall into the furnace, they would travel through the slag and into the molten metal, creating a pathway into the bath. Liquid water would follow in the pathway and molten metal would close in behind the water. Although ordinarily water in molten metal does not explode, the combination of splashing, impact, and high temperature would result in explosion. Belton based his opinion on a study in which water was sprayed from a hose onto molten aluminum. However, he was unable to give a definite opinion as to whether a brick or lance falling into the furnace would have sufficient impact to cause an explosion. He stated that the degree of impact required for such an explosion, is something "not well understood" and that the water and aluminum experiment on which he relied was "not very well designed" with regard to the impact factor.

Bieniosek, on the other hand, stated that a lance or brick falling onto a vapor and water layer in an open hearth furnace would not cause an explosion. According to Bieniosek, [*26] not only are impact and turbulence necessary under the circumstnaces described by Belton, but in addition the water must be trapped between the falling object and molten metal, rather than traveling behind. In any case, Bieniosek did not believe that falling bricks could penetrate the slag layer due to the highly viscous condition of the slag. Although a falling lance would penetrate the slag layer, Bieniosek believed that water pushed into the molten metal would have ample opportunity to escape around the pipe and into the furnance.

Even if viewed in the light most favorable to the Secretary, it cannot be said that Belton's testimony preponderates over Bieniosek's. At best the two experts' testimony is evenly balanced. Nor does other evidence in the record regarding the accumulation of large quantities of water in the furnaces tend to add weight to either expert's testimony. Several millwrights testified that liquid water can and does accumulate on the surface bath, but the Division Superintendent at Fairless testified that an accumulation of two to three inches is an unusual condition, and that no lance change would be ordered under unusual conditions. He also testified [*27] that large amounts of water would not accumulate in the furnace after millwrights go up on the roof because a newly leaking pipe or lance would be turned off at the source. This testimony was never rebutted.

The Secretary also introduced into evidence USS's "Safe Job Procedure for Changing Lances," and its "Safe Job Procedure for Lowering Jets After Hot Metal Addition." "Safe Job Procedures" are documents drafted by USS purporting to set forth safe procedures for performing various tasks. The "Safe Job Procedure for Changing Lances" provides, among other things, precautionary measures to be taken before and during a molten metal lance change. The precautions include: (1) no lance change during the first 60 minutes after the addition of hot iron to scrap and ore, or thereafter until the bath has been deemed stable by the melter foreman and first helper, (2) no lance change unless all driven fuel has been turned off 20 minutes, and (3) no pit or buggy activity on the charging floor or addition of ore to adjacent furnaces during the procedure. Item 1.4.1 of the "Safe Job Procedure for Lowering Jets After Hot Metal Addition" provides: "If the jet hangs up and causes slack in cable, [*28] the sudden drop can cause cable to break and jet will fall free. If hose pinches or breaks water leakage can cause roof damage and water leakage into furnace can cause explosion." The Secretary contended that these two safe job procedures, together with the historical rule against changing lances over molten metal, constitute an admission by USS that molten metal lance changes can cause explosions.

However, one of the millwrights who participated in formulating "Safe Job Procedure for Changing Lances" testified that it was instituted after an employee working over the roof was caught in a crane cable. He also stated that the purpose of the ban on molten metal lance changes was "for nobody to be on top of that roof," but inasmuch as the record establishes that millwrights do go up on the roof to repair door frames, even when there is hot metal in the furnace, we can infer very little from that statement. n9 Moreover, the Division Superintendent testified that at the time this Safety Rule was developed, furnaces were constructed of silica, rather than basic brick, and the dangers of roof collapse were considerably greater then. Finally, the "Safe Job Procedure for Changing Lances" [*29] makes no mention of an explosion hazard.

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n9 The judge found as a matter of fact that "there was an ironclad rule that, once hot metal from the blast furnace had been added to the open hearth furnace, no employees were allowed to go up on the roof superstructure of the open health furnace." However, the record contains uncontradicted testimony that millwrights do go up on furnaces containing molten metal in order to repair door frames.

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

The "Safe Job Procedure for Lowering Lances After Addition of Hot Metal," does refer to explosion hazards created by water leaks. However, as indicated earlier, the period immediately after addition of hot iron to the unmelted scrap and ore is known to be a particularly dangerous stage of the heat for reasons not related to water leaks. More importantly, it is clear from the record that the terms "explosion," "violent reaction," and "turbulence" are used interchangeably my millwrights. Consequently, use of the term "explosion" on this Safe Job Procedure does not constitute a concession [*30] that entrapment occurs; it may merely refer to a matter not at issue in the case, i.e., turbulence caused by water leaks in the absence of entrapment.

For these reasons we conclude that the record developed below does not demonstrate by a preponderance of the evidence that USS failed to render its workplace free of the alleged hazard of explosions resulting from entrapment of water below molten metal caused by changing oxygen lances above a furnace containing molten metal. Accordingly, the citation is vacated.

SO ORDERED.

CONCURBY: COTTINE

CONCUR:

COTTINE, Commissioner, concurring:

I concur in my colleagues' conclusion that the Secretary failed to establish that United States Steel Corporation ("USS") violated section 5(a)(1) of the Act, the "general duty" clause. n1 However, I adhere to the Commission's view in Bomac Drilling, Div. of TRG Drilling Corp., 81 OSHRC 45/A2, 9 BNA OSHC 1681, 1981 CCH OSHD P25,363 (No. 76-450, 1981), and its framework for analyzing alleged section 5(a)(1) violations. Accordingly, I would not overrule Bomac. However, the citation in this case should be vacated because the Secretary failed to establish that the occurrence of an incident was reasonably foreseeable. [*31]

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n1 Section 5(a)(1), 29 U.S.C. § 654(a)(1), is quoted at note 2 of the lead opinion.

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

The analysis in Bomac was based on the Commission's extensive experience in dealing with section 5(a)(1) cases. This experience demonstrated that many section 5(a)(1) citations were issued following specific incidents that led to death or injury in the cited workplaces. In these situations, the parties tended to focus on the specific incidents rather than on the general hazards alleged in the citations. This Commission and reviewing courts repeatedly pointed out that the proper inquiry was whether the general hazard was recognized within the meaning of section 5(a)(1), not whether a particular incident that had resulted in death or injury constituted a violation of section 5(a)(1). See General Dynamics Corp., Quincy Shipbuilding Div. v. OSHRC, 599 F.2d 453 (1st Cir. 1979); Brennan v. Butler Lime and Cement Co., 520 F.2d 1011 (7th Cir. 1975); Brennan v. OSHRC (Vy Lactos Laboratories, Inc.), 494 F.2d 460 (8th [*32] Cir. 1974); Western Massachusetts Electric Co., 81 OSAHRC 63/B13, 9 BNA OSHC P25,470 (No. 76-1174, 1981); Beaird-Poulan, a Div. of Emerson Electric Co., 79 OSAHRC 21/D11, 7 BNA OSHC 1225, 1979 CCH OSHD P23,493 (No. 12600, 1979); Boeing Co., Wichita Div., 77 OSAHRC 188/D13, 5 BNA OSHC 2014, 1977-78 CCH OSHD P22,266 (No. 12879, 1977). Despite these admonitions, the parties continued to focus on the foreseeability of specific incidents rather than on general hazards in many cases.

In Bomac, as in previous cases, n2 the Commission stressed that hazards in section 5(a)(1) cases should be defined generically, in order to avoid the tendency to focus on a specific incident. However, inquiring whether the employer recognized the generic hazard does not take into consideration whether the employer had the ability to prevent the occurrence of an incident that could endanger its employees. Because an employer's duty under section 5(a)(1) is an achievable one, a section 5(a)(1) violation can only be found if the employer had the ability to prevent the occurrence of such an incident. See National Realty and Construction Co. v. OSHRC, 489 F.2d 1257, 1266 (D.C. Cir. 1973). [*33] If an employer, in the exercise of reasonable diligence, cannot foresee that a hazardous incident will occur, the employer cannot be found in violation of the Act. See General Electric Co., 81 OSAHRC 42/A2, 9 BNA OSHC 1722, 1733, 1981 CCH OSHD P25,345 at p. 31,455-5 (No. 13732, 1981) (Cottine, Commissioner, concurring in part and dissenting in part). As stated by the U.S. Court of Appeals for the First Circuit, "An employer is not an insurer and need not take steps to prevent hazards which are not generally foreseeable . . . but at the same time an employer must do all it reasonably can to prevent foreseeable hazards." General Dynamics Corp., 599 F.2d at 458. Therefore, requiring that an incident be reasonably foreseeable is necessary to assure that the employer is not held to a standard of strict liability under section 5(a)(1). See National Realty & Constr. Co. v. OSHRC, supra.

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n2 See, e.g., Brown & Root, Inc., 80 OSAHRC 97/A2, 8 BNA OSHC 2140, 1980 CCH OSHD P24,853 (No. 76-1296, 1980).

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - [*34] - - - - - - -

Bomac presents a coherent, integrated approach to the adjudication of section 5(a)(1) cases. The generic definition of hazards assures that the citation and the Secretary's evidence focus on the general hazard the citation seeks to abate, rather than on a specific incident that may or may not be foreseeable. Requiring proof that an incident associated with the generic hazard is reasonably foreseeable assures that the duty imposed on the employer under section 5(a)(1) is achievable. n3

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n3 As the lead opinion points out, the Secretary must also prove that an incident is likely to cause death or serious physical harm and that a feasible means exists to eliminate or materially reduce the hazard in order to establish a section 5(a)(1) violation.

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

In this case, the generic hazard is explosions caused by the entrapment of water in molten metal. USS concedes that it recognizes this hazard, and previous cases have shown that the steel industry generally recognizes the hazard. See Babcock & Wilcox [*35] Co. v. OSHRC, 622 F.2d 1160, 1164 (3d Cir. 1980); Empire Detroit Steel Div., Detroit Steel Corp. v. OSHRC, 579 F.2d 378, 383 (6th Cir. 1978). Moreover, the work-site was not free of the hazard because of the unavoidable presence of water in USS's open hearth furnaces necessitated by the use of water as a coolant in the steelmaking process. However, the issue tried by the parties was whether an entrapment explosion could foreseeably occur as a result of an oxygen lance change over a furnace containing molten metal. In my opinion, the Secretary failed to prove that this incident was reasonably foreseeable.

The Secretary's expert witness, Mr. Belton, postulated two methods by which a lance change could result in an entrapment explosion. First, he suggested that the lance itself, which was filled with water, could fall into the furnace, the ports through which water entered and left the lance could become blocked with solidified slag, and the water thereby entrapped could cause the lance to explode. However, USS's expert, Bieniosek, believed slag would not solidify in the water ports because the water would immediately turn to steam when the lance fell into the molten metal [*36] and the steam rushing out o the ports would prevent slag from solidifying and water from becoming entrapped. Both experts were well-qualified, but Belton was not personally familiar with USS's furnaces, and his opinion was based on two assumptions not shown to be present. He assumed that the slag depth was eight to twelve inches when it was in fact three to four inches, and he assumed that the ports were one and one-half inches in diameter when they were in fact two inches in diameter. Thus, his opinion was based on the presence of more slag than was actually present and on the belief that it would become solidified in smaller openings than the lances contained. These important variations diminish the weight to be assigned Belton's opinion. Considering the record as a whole, the Secretary has not proven it was reasonably foreseeable that a lance falling into a furnace could lead to an entrapment explosion.

Belton next suggested that if a large quantity of water entered the molten metal bath, as could occur from a leak in the water pipes in the roof, and if a solid object then fell into the bath, the impact of the solid object could drive some of the water into the molten metal [*37] and an explosion could result. However, this type of explosion could only occur if the impact of the falling object is sufficient to entrap water between the molten metal and the falling object, and Belton was uncertain whether a falling brick or lance would have sufficient impact to do this. Bieniosek, on the other hand, was of the opinion that neither bricks falling from the furnace ceiling nor a falling lance could produce the necessary impact. Under these circumstances, I conclude that the Secretary failed to prove that this type of explosion was reasonably foreseeable.