PER CURIAM:

The Chlorine Institute, Inc. petitions for review of an order of the Occupational Safety and Health Administration (OSHA). By order of this court on October 9, 1979, petitioners' motion to stay the OSHA order was granted. Today we remove the stay and reinstate the OSHA order.

This case abounds with examples of bureaucratic ineptness. When the Occupational Safety and Health Act was passed in late 1970, Congress provided that the Secretary of Labor could, for two years from the effective date of April 28, 1971, circumvent the normal rule-making procedures and promulgate as an OSHA standard any "national consensus standard" and "any established Federal standard." 29 U.S.C. § 655(a) (1976). On the effective date, two federal standards existed on occupational exposure to chlorine gas. The first of the two standards was the "Threshold Limit Values of Airborne Contaminants for 1968" of the American Conference of Governmental Industrial Hygienists (ACGIH), which was printed at 41 C.F.R. § 50-204 (1968 standards). The ACGIH standard, adopted by the Secretary of Labor in 1969, included a table listing various gases, vapors, and fumes along with corresponding permissible concentration levels. If the contaminant was preceded by the letter "C," it indicated that the limit was a "ceiling" or maximum exposure amount. Absence of a "C" meant the exposure amount was computed over an eight-hour period as a time-weighted average (TWA). A TWA permits excursions above the limit so long as they are compensated for by corresponding periods below the limit. The 1968 table limited chlorine exposure to one part of chlorine per million parts of air (1 ppm) measured as a ceiling, not as an eight-hour average. This standard applied to industries engaged in the manufacture or production of chlorine products.

The second federal standard on chlorine exposure available to the Secretary in implementing the OHS Act was one concerning construction workers. One day before the OSH Act's effective date, the Secretary's regulations pertaining to the construction industry under the Contract Work Hours and Safety Standards Act (Construction Safety Act) became effective. These regulations adopted the ACGIH guidelines for 1970 (1970 standards). 29 C.F.R. § 1518.55 (now codified at 29 C.F.R. § 1926). On the 1970 tables, chlorine was not preceded by the letter "C," therefore, a time-weighted average was in effect.

Because ACGIH standards did not fall within the definition of "national consensus standard" under the OSH Act, 29 U.S.C. § 652(9) (1976), a ACGIH standard could only be adopted as an "established Federal standard." Id. § 655(a). The two existing federal adoptions of ACGIH standards (the 1968 and 1970 tables) conflicted. The 1968 standard, was relatively strict and allowed only a ceiling amount of 1 ppm chlorine. 41 C.F.R. § 50-204. The more liberal 1970 standard permitted a time-weighted average of chlorine over eight hours. 29 C.F.R. § 1518 (now 29 C.F.R. § 1926).

In choosing between the two conflicting federal standards the Secretary had unequivocal instructions from Congress.

In the event of conflict among any such standards, the Secretary shall promulgate the standard which assures the greatest protection of the safety or health of the affected employees.

29 U.S.C. § 655(a) (1976). Despite the clear mandate of the Act, the record is far from clear on which existing federal standard was adopted as the general contaminant level for chlorine. The standard promulgated by OSHA makes reference to both the 1968 and the 1970 ACGIH standards. At 29 C.F.R. § 1910.93, captioned "Air contaminants," the regulation stated it was incorporating the liberal 1970 standards. The table following the section was indeed the 1970 table of permissible exposure levels. The chlorine listing gave a time-weighted average and was identical to the 1970 ACGIH standard previously adopted at 29 C.F.R. § 1518 (now codified at 29 C.F.R. § 1926). Viewed by itself, section 1910.93 clearly adopted the 1970 standard. Problems were created, however, by 29 C.F.R. § 1910.99, the "Sources of standards" section. Section 1910.99 cited the stricter 1968 ACGIH standard, 41 C.F.R. § 50-204.50, as the source for section 1910.93. Obviously a mistake had been made either in section 1910.93 or 1910.99.

Evidently this was not the only mistake the government made. Two and one-half months later, the Secretary published "Miscellaneous Amendments" to the regulation. 36 Fed.Reg. 15,101 (1971). The Secretary stated the public had "pointed up the need for clarifications, corrections, and changes in effective dates of standards." As to section 1910.93, the Secretary stated it had been "revised in its entirety, in the interest of greater intelligibility and accuracy." Section 1910.99 (Sources of standards) was not mentioned. Section 1910.93 was amended to include an explanation of the "C" notation system. Reference to the 1970 standards was removed and many of the values in the table following the section were amended to reflect the 1968 standards. The "C" notations that preceded three entries under the 1968 standards, however, were not added to the regulation. Chlorine was one of these entries. The asterisk noting a 1970 change remained on the table before chlorine. Therefore, although some changes had been made to align the regulation with the stricter 1968 standards, the chlorine entry remained at its looser, 1970 time-weighted average amount.

The regulation was republished again in October, 1972. 37 Fed.Reg. 22,102 (1972). The purpose of the revision was to incorporate changes and correct "typographical and clerical errors." Section 1910.93 again included the 1970 standard for chlorine; section 1910.99 still listed the 1968 standard as the source of section 1910.93.

The two-year limit for passing regulations without resort to normal rule-making procedures passed. The Secretary cited employers for violating the time-weighted average, or 1970, standard. Secretary of Labor v. FMC Corporation, 2 OSHC 3216 (1974). In 1976, the National Institute for Occupational Safety and Health, an organization established by the OSH Act, recognized the 1970 1 ppm, TWA standard in a "Criteria Document" and suggested that the standard be changed to a 0.5 ppm, ceiling. On December 8, 1978, seven and onehalf years after it promulgated the regulation, the Secretary announced that "[t]he notation 'C,' designating a ceiling exposure limit, was inadvertently omitted" from the chlorine entry and two others. 43 Fed.Reg. 57,601 (1978). The Secretary stated that section 1910.93 (now section 1910.1000) incorporated the 1968 standards by reference, although actually the 1970 standards had been listed in the original section 1910.93. Id. at 57,602. The Secretary stated that because the "corrections" did not "establish, modify or revoke substantive rights and obligations," public notice and comment, and normal regulation-making procedures were unnecessary.

After the Chlorine Institute filed comments on February 1, 1979, OSHA issued a stay of the December 8, 1978 "corrections." Deciding that the tables after section 1910.93 were merely "to make it easier for employers and employees to have quick access" to the information and that 1910.99 (now 1910.1499) incorporated the 1968 standards, OSHA lifted its stay on July 17, 1979, and made the effective date of the "corrections" October 15, 1979. The Chlorine Institute petitioned this court for review and petitioned OSHA for a stay of its order. OSHA denied the petition on September 13, 1979. R. 444-48. In denying the stay, OSHA cited its conclusion that the 1968 standards provide the greatest protection for employees. Id. at 445. The court stayed the effectiveness of the corrections on October 9, 1979, pending review of the OSHA order.

The Secretary argues that omission of the "C" from the table was merely a ministerial mistake, which the Secretary can now correct without resort to its regulation-making procedures. The courts have long recognized that administrative agencies may correct inadvertent, ministerial errors. American Trucking Ass'ns, Inc. v. Frisco Transportation, Co., 358 U.S. 133, 79 S.Ct. 170, 3 L.Ed.2d 172 (1958) (ICC certificate); United States v. Seatrain Lines, Inc., 329 U.S. 424, 67 S.Ct. 435, 91 L.Ed. 396 (1947) (ICC certificate); Bell v. Hearne, 60 U.S. (19 How.) 252, 15 L.Ed. 614 (1857) (title to land). Not only must the error be inadvertent and ministerial, but the power to correct "may not be used as a guise for changing previous decisions because the wisdom of those decisions appears doubtful in the light of changing policies." American Trucking Ass'ns v. Frisco Transportation Co., 358 U.S. at 146, 79 S.Ct. at 177.

Much commends the Chlorine Institute's position that the December 8, 1978 announcement reflects not a ministerial "correction," but a change in policy. Seven and one-half years elapsed before this clerical error earned administrative attention. Although OSHA argues it was the Secretary's intent to adopt the 1968 standards from the start, in as early as 1974 OSHA went to the trouble of citing employers under the more liberal 1970 standards without beginning the mechanisms for either correcting the error or modifying the regulation. The Institute argues it strains credulity to claim that OSHA erred not only in the original promulgation of the regulation, but also in the subsequent amendments and corrections of it. The Institute points to the history of the use of "ceiling" as opposed to time-weighted averages for chlorine exposure as proof that the Secretary chose 1970 standards, because the 1968 ceiling standard was an anomaly which was quickly changed. The petitioner also asserts that the eleventh hour promulgation of the 1970 standards under the Construction Safety Act indicates the Secretary's intent to utilize the 1970 standards as the "established Federal standards" in this regulation. In essence, the Institute argues the ministerial error was not the omission of "C" from the table, but the incorrect listing of the 1968 standard (41 C.F.R. § 50-204) as the source of 29 C.F.R. § 1910.93.

Although difficult, it is not impossible to believe OSHA took over seven years to uncover an error such as this. The chlorine standard is one of nearly 300 adopted by OSHA in 1971. Between thirty and forty discrepancies were corrected. While the 1978 correction could appear to be a policy change prompted by the NIOSH suggestions, the chlorine listing is not the only error being corrected: the "C" was omitted on two others, and one substance was completely omitted. Because the 1968 standard was stricter, the error was not one which employers would raise when cited for violations of the 1970 standard. The Secretary argues that no conflict really exists in the standards; the liberal 1970 standard applied to construction workers and the stricter 1968 standards to manufacturing workers. The timely adoption of the 1970 standards does not indicate an intent to incorporate them in section 1910.93. Those standards were included in the regulations, but only for the construction industry. 29 C.F.R. § 1910.12. The Secretary asserts that even if a conflict exists between the standards, OSHA is required to "promulgate the standard which assures the greatest protection." 29 U.S.C. § 655(a) (1976). OSHA has found that ceiling limits provide greater protection than TWAs and the record discloses substantial evidence in support of this conclusion. Finally, the Secretary points to the sources section as clear evidence of an intent to adopt the 1968 standards.

Clearly a clerical mistake exists in one of the sections. While we certainly do not condone OSHA's inattentiveness and tardiness, substantial evidence exists to show that the error is in 1910.93, not 1910.99. For the Secretary to have done otherwise would have been a transgression of the precise instructions given by Congress. We therefore remove the stay and uphold the Secretary's actions.

AFFIRMED.