Full opinion text
FINDINGS OF FACT, CONCLUSIONS OF LAW, AND ORDER FONG, District Judge. The above-entitled case came on for trial before this court on January 5, 1993. Plaintiffs Hawaii’s Thousand Friends and Sierra Club (“plaintiffs”) appeared through their attorneys Sierra Club Legal Defense Fund, Inc., Paul P. Spaulding, III, Denise E. Antolini and Eric S. Walters. Defendant City and County of Honolulu (“the city” or “defendants”) appeared through Deputy Corporation Counsel Cheryl K. Okuma-Sepe and Tracy Lowell Wolf. On March 27, 1990, plaintiffs filed a complaint for injunctive relief and civil penalties against the city arising out of allegedly illegal bypasses of sewage treatment equipment at the city’s Honouliuli wastewater treatment plant (“Honouliuli I ”). On July 3, 1991, this court entered an order in which it granted plaintiffs’ motion for partial summary judgment on the issues of standing, subject matter jurisdiction, and the city’s liability for 104 violations of the Clean Water Act: 52 illegal bypasses and 52 failures to report those bypasses. On December 24, 1991, plaintiffs filed a second complaint for declaratory judgment, injunctive relief and civil penalties against the city, which also relates to the city’s operation of its Honouliuli plant (“Honouliuli II ”). Specifically, plaintiffs alleged that the city has violated the Clean Water Act on a continuous basis since July 1, 1988 by failing to treat sewage at secondary levels. On February 28, 1992, this case was consolidated with the earlier lawsuit. On May 8,1992, the court granted plaintiffs’ motion for summary judgment on liability, finding that the city has violated the secondary treatment requirements of both the Act and the Honouliuli National Pollution Discharge Elimination System (“NPDES”) permit incorporating the Act’s requirements on a daily, continuous basis since July 1, 1988. The purposes of this trial are to establish the precise number of secondary treatment violations, to examine what remedies are appropriate to address both the bypass and secondary treatment violations, and to determine appropriate civil penalties pursuant to 33 U.S.C. § 1319(d). This court, having examined the documentary and other evidence introduced, heard the oral testimony, considered the arguments of counsel, and reviewed the written memoranda of the parties, makes the following findings of fact and conclusions of law. FINDINGS OF FACT I.BACKGROUND A. The Honouliuli Plant 1. This lawsuit concerns the Honouliuli wastewater treatment plant (“the Honouliuli plant” or “the plant”), which is a sewage treatment facility owned and operated by the City and County of Honolulu (“the city”), and originally designed to treat 25 million gallons per day (“mgd”) of sewage. It is one of twelve sewage treatment plants on Oahu operated by the city. 2. The plant is located in Ewa Beach, Honolulu, Oahu, Hawaii, and treats wastewater from the southwestern Oahu metropolitan area, ranging from Halawa to Makakilo. The plant’s service area is predominantly residential and agricultural in character, with some commercial areas. 3. The Honouliuli plant currently is a “primary” sewage treatment facility. After sewage is pumped to the plant through the collection system, it receives preliminary treatment consisting of screening to remove large objects, grit removal and pre-aeration to facilitate grease and floatables removal. The sewage then enters the primary clarifiers, which are designed to remove suspended solids and organic materials through sedimentation to the bottom of the tank and by removal of floatables from the surface. The sewage then travels through an effluent screen (with a inch square mesh) and leaves through the outfall pipe. The effluent sewage is dispersed into the ocean waters through a multiport diffuser. 4. The plant currently processes an average of approximately 23-25 mgd. The treated sewage is then discharged through an ocean outfall into Mamala Bay, offshore of Ewa Beach and adjacent to civilian and military beaches from Iroquois Point to Barber’s Point. 5. The Honouliuli plant is the second largest municipal wastewater discharge facility in the State of Hawaii. B. Parties 6. Plaintiff Hawaii’s Thousand Friends is a non-profit corporation concerned with water and land use, organized under the laws of the State of Hawaii. 7. Plaintiff Sierra Club is a national nonprofit conservation organization. It has a Hawaii chapter that currently has more than 4,000 members. 8. In its July 3, 1991 order, the court found that some members of Hawaii’s Thousand Friends and Sierra Club reside near and use the Ewa beaches and coastal waters as more precisely explained therein. Accordingly, the court held that plaintiffs satisfy the organizational and personal standing requirements for maintaining these citizen lawsuits under the Clean Water Act. 9. Defendant City and County of Honolulu (“the city”) is a Hawaii municipal corporation that owns and operates the Honouliuli plant. The Wastewater Management Division (“WWMD”) of the Department of Public Works is the city division that oversees operation of the Honouliuli Plant. One branch of the WWMD, the Wastewater Treatment and Disposal branch (“WWT & D”), is responsible for the administration of the city’s treatment plants. C. Notice of Citizens’ Suit 10. On November 28, 1989, plaintiffs served written notice on the city, the United States Environmental Protection Agency (“EPA”) and the State of Hawaii Department of Health (“DOH”) of their intent to file a citizen’s lawsuit on the claims set forth in the complaint in Hawaii’s Thousand Friends v. City and County of Honolulu, Civil No. 90-00218 HMF (‘Honouliuli I ”). 11. This written notice is required by section 505(b)(1)(A) of the Clean Water Act, 33 U.S.C. § 1365(b)(1)(A). 12. The November 28,1989 written notice was received by the city, EPA and DOH. 13. After November 28, 1989, neither EPA nor DOH commenced and diligently prosecuted a court action to address the violations in this notice. 14. Plaintiffs filed their complaint in Honouliuli I on March 27, 1990, more than sixty days after service of the November 28, 1989 notice. 15. As explained in the court’s September 28, 1990 and July 3, 1991 orders, the allegations set forth in the Honouliuli I complaint satisfied the jurisdictional doctrine of Gwaltney of Smithfield v. Chesapeake Bay Foundation, 484 U.S. 49, 108 S.Ct. 376, 98 L.Ed.2d 306 (1987) (“the Gwaltney doctrine”). 16. On October 16, 1991, plaintiffs served written notice on the city, EPA, and DOH of their intent to file a citizens’ lawsuit on the claims set forth in the complaint in Hawaii’s Thousand Friends v. City and County of Honolulu, Civil No. 91-00739 ACK (“Honouliuli II”). 17. The October 16, 1991 written notice was received by the city, EPA and DOH. 18. After October 16, 1991, neither EPA nor DOH commenced and diligently prosecuted a court action to address the violations in the October 16, 1991 notice. 19. Plaintiffs filed their complaint in Honouliuli II on December 24,1991, more than sixty days after service of the October 16, 1991 notice. 20. The violations set forth in the Honouliuli II complaint have continued on a daily basis since October 16, 1991 and therefore satisfy the jurisdictional requirements of the Gwaltney doctrine. D. The Clean Water Act 21. In 1972, Congress amended the Clean Water Act, 33 U.S.C. § 1251, et seq. (“the Act”), to create the National Pollution Discharge Elimination System (“NPDES”), which requires a NPDES permit for any person to discharge pollutants into navigable waters of the United States. 33 U.S.C. § 1311(a). The program originally required all publicly owned treatment works (“POTWs”) to achieve effluent limitations based on secondary treatment by July 1, 1977. 22. Effluent limitations have been set by the Environmental Protection Agency pursuant to 33 U.S.C. § 1314(d)(1). The EPA defines “secondary treatment” to require the removal of 85% of BOD and SS in the effluent, average concentrations of 30 mg/1 on a 30-day average for both BOD and SS, average concentrations of 45 mg/1 on a seven-day average for both BOD and SS, and a pH level between 6.0 and 9.0. 40 C.F.R. §§ 133.102(a) and (b). 23. Sewage plant engineers generally refer to secondary treatment as the use of biological treatment (i.e., digestion by microorganisms) of sewage in addition to the physical treatment processes provided by primary treatment. 24. Primary treatment, the lowest level of wastewater treatment, is generally defined as the physical treatment of effluent through screening and gravity settling in sedimentation tanks. It is defined by the Clean Water Act as at least 30% removal of BOD and SS. 33 U.S.C. § 1311(h). Advanced primary treatment normally involves the addition of chemicals to enhance settling and coagulation of sewage matter and removes a higher percentage of BOD and SS than primary treatment. 25. The Act has been amended several times to extend the secondary treatment compliance deadline for POTWs to its current and final date of July 1, 1988. No extensions can be granted beyond that date by the EPA or a state. 33 U.S.C. §' 1311(i)(l). 26. In 1977, Congress amended the Act to allow municipalities that discharge into deep marine waters to obtain NPDES permits for discharge at less than secondary treatment effluent limitations if they could establish that it would not harm the marine environment. 33 U.S.C. § 1311(h). 27. On June 15, 1979, the EPA promulgated regulations governing § 301(h) applications. 44 Fed.Reg. 34,784-34,832. In response to NRDC v. USEPA, 656 F.2d 768 (D.C.Cir.1981), the 301(h) regulations were amended on November 26, 1982, to allow municipalities to submit a reapplication following a tentative decision to propose changes in treatment level or outfall and diffuser design. 40 C.F.R. Part 125.59(d). E. The State NPDES Program 28. Pursuant to 33 U.S.C. § 1342, the Administrator of the EPA has delegated to the State of Hawaii the power to issue certain National Pollution Discharge Elimination System (“NPDES”) permits to dischargers in Hawaii. DOH is the state government department that administers the NPDES program in Hawaii. 29. The state program for issuance of NPDES permits must comply in all respects with, and must prescribe no less stringent effluent limitations than, those set forth in the Act itself. 33 U.S.C. § 1342(b). 30. In issuing an NPDES permit, both DOH and EPA are required to prescribe effluent limitation conditions that meet the discharge requirements set forth in the Act. 33 U.S.C. § 1342(a) and (b). 31. All permits issued by the state must also comply with Section 308 of the Act, 33 U.S.C. § 1318, which requires NPDES permittees to establish and maintain records; install, use and maintain monitoring equipment; sample effluent; and report on a regular basis to DOH and EPA regarding their discharge of pollutants using discharge monitoring reports (“DMRs”). 33 U.S.C. § 1342(b)(2)(A). 32. The city has filed DMRs with DOH on a periodic basis since it first received an NPDES permit for the Honouliuli plant. In these DMRs, the city is required to record the daily, weekly, and monthly measurements required by its permit. F. Development of the Honouliuli Plant 33. In April 1970, the city engaged a consortium of engineering consultants to undertake a comprehensive study of the present and future sewage treatment needs of Oahu. The consortium, assisted by a six-member Board of Advisors, issued its final report in April 1972, which was entitled “Water Quality Program For Oahu With Special Emphasis On Waste Disposal” (the “WQPO report”). 34. The WQPO report found that sewage treatment on Oahu was inadequate in many areas (indeed, raw untreated sewage was being discharged into some receiving waters), resulting in a gradual deterioration of the water quality of some of the coastal waters receiving the effluent. According to the Report, some areas—including Mamala Bay, Pearl Harbor and Kaneohe Bay—-had already manifested this deterioration and required construction of new treatment facilities. 35. The WQPO report concluded that a wastewater treatment and disposal system was urgently needed and recommended an overall plan for water quality management on Oahu. Specifically, the report recommended that a secondary treatment disposal sewage facility be built at Honouliuli to handle waste flows from Halawa to the Ewa areas, “with primary consideration of reclamation for irrigation of sugar cane in the Ewa District.” WQPO report at 11-17. The report also contemplated that an ocean disposal system would be required at a later time to handle effluent in excess of the reuse need. The report suggested that because of the favorable current regime and density structure at reasonable distances from shore in Mamala Bay, “the minimum degree of treatment required for the Sand Island or Honouliuli systems (discharge into deep waters off Ewa Beach), is advanced primary, if reclamation is not considered.” Id. at XIII-15. 36. Consistent with the recommendation of the WQPO report, the city and its consultant, the R.M. Towill Corporation (“Towill”), with assistance from various state and federal agencies, planned a secondary treatment plant (using conventional activated sludge technology) at the Honouliuli site. 37. The land for the Honouliuli plant itself was acquired by the federal government in 1974. It was formerly a portion of Barber’s Point Naval Air Station. 38. In June 1975, the city issued a Final Environmental Impact Statement for the Honouliuli plant, which had been prepared by Towill. The “proposed action,” on which the environmental analyses were based, was the construction of “a secondary activated sludge wastewater treatment plant” and an ocean outfall system. 39. In August 1976, Towill, pursuant to a contract with the city, issued a final design report for a secondary treatment plant at Honouliuli. The city then proceeded to have detailed plans and specifications drawn for construction of this plant. 40. The WQPO report had also recommended that a Honouliuli outfall/diffuser system be built to discharge treated sewage into West Mamala Bay. In 1976, to implement this recommendation, the city and Towill issued a design report for construction of this outfall (the “outfall design report”). 41. In the outfall design report, Towill proposed an outfall system that consisted of 9,166 feet of 84-inch outfall pipe on land and 8,760 feet of 78-inch outfall pipe along the ocean bottom. The outfall terminated in a 1750-foot diffuser at a depth of approximately 200 feet. The diffuser had a graduated pipe diameter, decreasing to 48 inches at the end in order to maintain adequate velocity of the effluent flowing through the pipe. The diffuser had uniformly spaced ports through which the effluent would be expelled into the receiving waters by pressure of its flow. 42. From approximately 1977 to 1980, the city built the outfall recommended by Towill (with 84-inch and 78-inch pipe) to discharge up to 112 mgd of treated sewage effluent. 43. In the late 1970s, after completing preparations to build a secondary treatment plant at Honouliuli, the city decided to apply for a waiver of the secondary treatment requirement pursuant to the newly enacted § 301(h) of the Act (the “301(h) waiver”). The city submitted an application for an NPDES permit to build an “advanced primary” plant on September 7, 1979. The “advanced primary” treatment was to be accomplished through use of a polymer (chemical) feed system to enhance the pollutant removal rates provided by primary treatment equipment. 44. On October 31, 1980, EPA issued an NPDES permit to the city that governed the operation of the Honouliuli plant until June 30, 1985. This permit imposed secondary treatment effluent limitations (85% removal rates and 30 mg/1 concentration limitations) on the plant for suspended solids and organic materials in the effluent. 45. On March 6, 1981, EPA notified the city that it had deferred consideration of the waiver applications for the Kailua, Kaneohe and Waianae sewage plants. EPA stated that it would not require secondary treatment at those plants pending review of the waiver applications. EPA further explained that any NPDES permit or § 309(a)(5)(A) extension orders setting interim effluent limitations (based on past performance and existing plant capabilities) would remain in effect until the waiver was granted or denied. This letter did not include the waiver application for the Honouliuli plant as one of the deferred applications. 46. In September 1981, the EPA tentatively denied the city’s 301(h) waiver proposed for suspended solids (with effluent limitation of 115 mg/1) and tentatively granted a waiver for BOD materials (200 mg/1). 47. The city had new plans and specifications drawn for a primary treatment plant and began building it in 1982. The plant was completed and began treating sewage in December 1984. In doing so, the city chose to take the chance that its 301(h) waiver would be granted and thus did not build the more expensive secondary treatment facility which it would not be required to have if and when the 301(h) waiver is ultimately granted. 48. The city has recently completed a 13 mgd expansion of the Honouliuli plant. The plant now has a total of four primary clarifiers—three of which are generally in use at a given time—and can process an average flow of 38 mgd. 49. The area serviced by the Honouliuli plant is one of the fastest growing areas on Oahu. According to Dr. John Lewin, Director of DOH, it is “the area of planned major growth on the Island of Oahu.” There are numerous residential developments, which include golf courses and commercial districts, being planned for and built in this area. 50. As the population of the Honouliuli service area grows, the flow of pollutants to the Honouliuli Sewage plant will grow in a corresponding manner. 51. The most recent “optimistic” city projections indicate that the sewage flows into the plant will grow from their current level of 23-25 mgd to approximately 42 mgd by 1995 and 51 mgd by the year 2000. These projections include development plans that have already been accepted by the city, as well as plans presently proposed by the developers. More conservative city projections based only on currently approved development plans indicate that sewage flows will reach only approximately 31 mgd by 1995, and 33 mgd by 2000. 52. The planned major growth in the Ewa plain area has also raised concerns about water needs in the area. Dr. Lewin testified that it would be logical to conserve water on the Ewa plain by reusing the waste-water treated at the Honouliuli plant as irrigation water and/or to recharge the aquifer. Dr. Lewin testified that secondary treatment, coupled with some disinfection process, is the sewage treatment level necessary to pursue reuse or recharge of the aquifer. 53. The city has begun planning for a secondary treatment facility capable of processing 13 mgd for effluent reuse. Construction of this facility is scheduled to be completed in 1996. The Board of Water Supply has projected that 12.5 mgd of the effluent can be reused. 54. The city plans to treat at secondary levels only the amount of effluent that will be reused. For example, if the reuse demand is only for 5 mgd, then only 5 mgd will receive secondary treatment. G. Required Effluent Limitations for the Honoidiuli Plant 1. Interim Effluent Limitations 55. In January 1982, the Honouliuli outfall began accepting sewage flows, as ordered by the State, that were diverted from Pearl Harbor. The partially completed plant discharged effluent that had only received preliminary treatment (screening and degritting). Over the next two years, eleven other sewage flows were diverted into this outfall before the treatment plant itself was completed. 56. On February 19, 1982, DOH issued the city a Notice and Finding of Violation for the plant’s failure to provide secondary treatment. On December 15, 1982, DOH issued the city an administrative order requiring the city to complete construction of the primary treatment units by December 31, 1982, and of the solids handling units by December 31, 1984. The order also levied a fine of $100,-000 against the city for the continuous discharge, payment of which was suspended as long as construction was completed on time. 57. On January 30, 1985, DOH issued an amendment to the December 1982 order (the “1985 consent order”). The 1985 consent order established interim effluent limitations of 200 mg/1 BOD and 105 mg/1 SS for the Honouliuli plant, as proposed by the city in the municipal compliance plan it submitted to DOH. It also established a construction schedule for the completion of secondary treatment facilities, to take effect upon the denial of the city’s 301(h) waiver application. 58. The interim effluent limitations and construction schedule were developed with EPA Region IX approval and directions from EPA headquarters. EPA and DOH intended the consent order to be valid after July 1, 1988. At the time the consent order was drafted, it was EPA Region IX policy to issue these orders administratively rather than through judicial means. 2. 1985 NPDES Permit 59. On July 1, 1985, DOH issued NPDES permit no. HI 0020877 (“the 1985 permit”) to the city for the Honouliuli plant. The permit recites that it will expire by its terms at midnight on June 30, 1990. 60. The permit authorized the plant to discharge wastewater into Mamala Bay from a sewage outfall, known as the “Barber’s Point outfall,” so long as certain specified conditions were met. 61. The effluent limitations contained in the plant’s NPDES permit are pegged to the secondary treatment requirements of the Clean Water Act. The permit does not authorize the discharge of primary or advanced primary sewage effluent. 62. The permit contains ten separate effluent limitations at issue in this case: five each for biochemical oxygen demanding (“BOD”) materials and for suspended solids (“SS”). 63. BOD materials are organic substances in wastewater effluent that bind oxygen, thereby depleting oxygen in water and degrading water quality. 64. Suspended solids are solid particulates contained in wastewater effluent. 65. The permit prohibits the plant from discharging more than an average of 6,248 pounds per day of BOD materials or SS over a 30-day period (limitation nos. 1 and 2), or more than an average of 9,371 pounds per day of BOD or SS in any seven-day period (limitation nos. 3 and 4). 66. The permit also prohibits the plant from discharging effluent with concentrations of BOD materials or SS that exceed an average concentration of 30 milligrams/liter per day over a 30-day period (limitation nos. 5 and 6) or more than an average concentration of 45 milligrams/liter per day in any seven-day period (limitation nos. 7 and 8). 67. The permit also imposes limitations on the “removal efficiency performance” of the plant for both BOD and SS. It requires that the plant remove 85% of all BOD materials and SS entering the plant (limitation nos. 9 and 10). 68. As the city admits, the Honouliuli plant has never met, and is not designed to meet, the secondary treatment effluent limitations contained in its NPDES permit. 69. The city has operated the Honouliuli plant on the assumption that the only effluent limitations the plant must meet are the interim limitations established in the 1985 consent order. 70. Despite the apparent belief by the city, the DOH, and the EPA that the effluent limitations established in the 1985 consent order were effective after July 1, 1988, the court has previously held that the July 1, 1988 compliance deadline could not be extended administratively. July 3, 1991 Order, at 21-23; see 33 U.S.C. § 1311(i)(l). H. 801 (h) Waiver Process for the Honouliuli Plant 71. On or about September 7, 1979, the city filed an application for a 301(h) waiver permit from the EPA. 72. On or about September 8, 1981, EPA tentatively denied the portion of the city’s 301(h) waiver request relating to a variance for suspended solids. EPA tentatively granted the portion of the city’s request for a waiver from BOD secondary treatment requirements. 73. On October 31, 1983, the city reapplied to EPA for a waiver of secondary treatment requirements at Honouliuli, as permitted under the amended regulations. By this application, the city downgraded its requested plant to “primary” treatment, rather than its original waiver proposal of “advanced primary” treatment. 74. The EPA tentatively granted the amended waiver for both BOD and SS on April 4, 1988. 75. Two public hearings were held to receive testimony on the tentative grant of the 301(h) waiver for primary treatment. On May 2, 1991, the EPA issued a “final determination” granting the city a 301(h) waiver permit. 76. On June 5, 1991, Sierra Club, Hawaii’s Thousand Friends and Save Ewa Beach Ohana filed an evidentiary hearing request on the waiver permit decision. 77. On June 5, 1991, the city also filed an evidentiary hearing request on certain portions of the decision. 78. These evidentiary hearing requests automatically stayed the effectiveness of the proposed 301(h) permit. 40 C.F.R. § 124.-15(b)(2). Thus, the proposed waiver never became final and effective. 79. On May 12, 1992, the city filed an emergency petition for writ of mandamus in the Court of Appeals for the Ninth Circuit, requesting that the court direct EPA to act expeditiously on the evidentiary requests filed by plaintiffs and by the city. This petition was filed four days after this court held that the city has been in continuous violation of the secondary treatment requirement. 80. On June 9, 1992, one month after the Court entered its summary judgment order finding secondary treatment violations in this case, the city withdrew its evidentiary hearing request. 81. On July 31, 1992, EPA granted plaintiffs’ evidentiary hearing request in full. An Administrative Law Judge (“ALJ”) was assigned on August 19, 1992. The ALJ requested that the city, the EPA, and plaintiffs submit position statements and suggested hearing dates by January 15,1993. The ALJ will now conduct hearings on a wide range of waiver issues. This is a de novo proceeding in which the city will carry the burden of proof to demonstrate its entitlement to a waiver. This hearing process, with associated appeals, may not be completed until 1995 or 1996. 82. While the 301(h) evidentiary hearing requests are pending before EPA, the proposed waiver permit is not effective. Accordingly, the city does not now have authority to discharge effluent at the treatment levels set forth in the proposed waiver permit. II. BYPASS INCIDENTS A. Regulations and Reporting Procedures for Bypasses at Honouliuli 83. Among other things, the permit specifically prohibits any “bypass” of sewage by the plant. Permit, Section II.A.lO.d. A “bypass” is defined as “the intentional diversion of waste streams from any portion of a treatment facility.” Permit, Section II.A.10. a(l). 84. Sewage flow into a wastewater plant that is diverted to the effluent channel after receiving only partial treatment and is sent through the outfall pipe to ocean waters constitutes a bypass. 85. Bypasses are allowed only if: (1) the bypass does not cause effluent limitations to be exceeded; and (2) the bypass is for “essential maintenance to assure efficient operation.” Permit, Section II.B.lO.b. These permitted bypasses are not subject to the notice requirements. 86. The permit requires the city to immediately report to DOH “any noncompliance which may endanger health or the environment.” Permit, Section II.A.2. In addition, the plant must give immediate notice to the major news wire services (UPI and AP) for broadcast so that the public can be alerted to the bypass event. Within five days, the plant must make a written report of the incident, which includes: a description of the noncompliance and its cause; the period of noncompliance, including the dates and times, and expected duration; and steps taken or planned to reduce, eliminate, and prevent recurrence of the noncompliance. 87. The permit also requires that, for an anticipated bypass, defendant must submit prior notice of said bypass to DOH, if possible, at least ten days before the date of the bypass. Permit, Section II.B.lO.c(l). 88. In the event of an unanticipated bypasses, notice must be given in the manner set forth in Section II.A.2 of the permit, as described above. Section II.A.2.a. specifies that any unanticipated bypass that exceeds any effluent limitation in the permit must be reported immediately. 89. The city is also obligated to comply with an internal directive of its WWT & D branch regarding sewage bypasses (the “directive”). This directive requires the Honouliuli plant to report immediately all emergency or anticipated “bypasses” or “overflows” of raw or partially treated sewage. These reports must be made verbally to someone in the chain of command (from branch superintendent to division chief to department director) in order to assure prompt notification of city officials, DOH, and the public. In addition, a wastewater bypass report form must be submitted to the branch superintendent on the first working day after each bypass occurrence. 90. The directive further requires that for a planned discharge of raw or partially treated sewage (i.e., overflows or bypasses due to repair, construction or testing), a wastewater bypass report must be submitted to the Water Quality section (a section of the WWMD) at least one month before the planned overflow or bypass. 91. The DOH investigation of the bypass incidents at Honouliuli concluded that the city did not follow the provisions of its own directive before or during the October-November 1989 bypass episodes. 92. The city is further obligated to comply with separate procedures promulgated by the director and chief engineer of the city’s Department of Public Works, contained in a document entitled “Standard Operating Procedure (SOP) for Reporting Bypass of Sewage or Discharge of Partially-Treated [sic] Sewage October 28, 1983 (Revised June 29, 1987).” This document also contains specific oral and written notification, media announcement, and posting procedures similar to those above for both emergency and anticipated bypasses. 93. All of the reporting and notification procedures set forth in the directive and the SOP are designed to protect the public and the ocean environment from the adverse effects of the discharge of inadequately treated sewage. Upon notification, DOH and the various city departments can implement measures to minimize or mitigate the bypasses, closely monitor the effluent and take other steps to warn the public (through posted signs or media announcements), and thereby significantly decrease the public health dangers and environmental impacts. Noncomplianee with the internal procedures does not constitute a violation of the NPDES permit or the Clean Water Act, but is relevant to whether the city acted in good faith. B. Bypasses at Honouliuli in October and November 1989 94. The Honouliuli plant is designed to process an average daily capacity of 25 million gallons of effluent. Athough the plant can handle occasional peak flows exceeding that amount, the city knew that the plant could bypass preliminarily treated sewage when there were sustained heavy flows, particularly if portions of plant treatment equipment were out of service. These bypasses could be expected to occur during periods of heavy rain and/or heavy usage (such as in the early evening or on weekends). Flows at the Honouliuli plant increase during heavy rains because of leaking sewer collection pipes and “manholes” that allow rain saturating the ground to flow into the sewer system. 95. According to plant employees, bypasses could occur at the plant, even with two primary clarifiers in operation, when flows exceeded 38 mgd. When only one primary clarifier was in operation, bypasses would occur when flows reached the 25.5 mgd level. 96. The Honouliuli plant periodically schedules maintenance of its wastewater treatment equipment, including the plant’s two primary clarifier units. When these repairs take place, the plant has a lower threshold level at which bypasses will take place because each primary clarifier treats an average of approximately 13 mgd. 97. On or about July 19, 1989, the city prepared an internal wastewater bypass report for an anticipated “bypass” of “partially treated wastewater” into Mamala Bay from the Honouliuli plant over a 30-45 day period starting on August 27, 1989. That internal report, prepared by Norman Ching, the plant superintendent (“Ching”), was circulated to and approved by numerous city WWMD employees. The report identified the cause of the bypass to be the “preventive maintenance” of primary clarifier # 1. The plan was to repair the deteriorated coating of the concrete and metal surfaces of clarifier # 1, which protects against corrosion. 98. On or before July 27, 1989, the notified DOH by letter that it planned to perform “scheduled preventive maintenance” for one of the two primary clarifiers for the plant. The city indicated that the remaining primary clarifier would be “hydraulically overload[ed,l ... which may decrease clarifier performance.” The city did not mention that it anticipated bypasses might occur as a result of the maintenance. The letter does not even use the word “bypass.” A hydraulic overload occurs when flows inside a primary clarifier rise above design levels, which may decrease detention time in the tank and reduce the efficiency of removal. According to DOH enforcement officer Paul Ciesla (“Ciesla”), a hydraulic overload of a clarifier is “vastly different” and “far less serious” than a bypass. 99. On or about August 4, 1989, the city received permission from DOH to perform this “scheduled preventative maintenance” during a period between August 27, 1989 through October 11, 1989. DOH acknowledged that the maintenance work would create a hydraulic overload on primary clarifier # 2, and urged the city to complete the work as expeditiously as possible. 100. After this notification to DOH, the city twice rescheduled the maintenance work on primary clarifier #1. On September 5, 1989, the city notified DOH that it had revised its schedule for maintenance to commence about six weeks later, on October 16, 1989 (instead of August 27, 1989). The city later moved this date up to October 3, 1989. 101. The city was aware that flows into the plant vary seasonally and that October is typically a month of high influent flows. 102. At no time did the city notify DOH that it anticipated that any bypasses would take place during the repair period. 103. On October 3, 1989, the city began its scheduled repairs, and clarifier # 1 was taken off-line. Bypasses of the plant’s remaining primary clarifier began occurring immediately—flows backed up and spilled from the “influent” channel (between the preaeration tanks and the primary clarifiers) over the “bypass weir” into the plant’s “effluent channel.” In other words, sewage that had received only preliminary treatment and was on its way to the primary clarifier spilled over into the sewage that was leaving the plant after treatment. The bypassing sewage skipped primary treatment altogether. The bypass weir was adjustable and designed to allow excess flow to enter the effluent channel directly, bypassing the clarifiers, in order to prevent hydraulic overload to the clarifiers. 104. Due to the heavy rainstorm flows that occurred on October 3, 1989, which contributed to the bypassing, the plant stopped the maintenance work on primary clarifier # 1 and put both clarifiers back on-line at approximately 8:45 p.m. 105. Primary clarifier # 1 was taken offline again on October 4, 1989, at about 12:30 p.m. Bypasses over the bypass weir continued on a daily basis until clarifier # 1 was placed back in service on November 23,1989. 106. A total of approximately 106 million gallons of preliminarily treated sewage was bypassed on 52 days during this period. The daily bypasses ranged as high in amount as 8.37 million gallons on October 3, 1989, and there were 23 separate days on which more than two million gallons of preliminarily treated sewage were bypassed. The bypass flows were the highest on weekends. 107. The plant was unable to put primary clarifier # 1 back in service quickly once the recoating work had begun and equipment was placed in the tank. The only measure undertaken by the plant to attempt to mitigate the bypasses was to initiate procedures to throttle back the pumps to keep the flow below 25 mgd whenever possible. This manipulation of the pumps was constrained by the fact that sewage could back up in the collection system or in the preliminary treatment equipment if the flows were overly restricted. The plant did not monitor the bypasses at the bypass weir itself. Plant employees, however, did take effluent samples four days a week to monitor the levels of BOD and SS; the levels never exceeded the interim effluent limitations, but were in excess of secondary treatment limitations. 108. The 106 million gallons of preliminarily treated sewage bypassed did not receive the complete primary treatment. The sewage was only degritted, screened, and aerated (or “conditioned”) prior to being bypassed into the plant’s effluent channel. The effluent screens did remove large floatables from the sewage, but the pre-aeration treatment had little effect on the water quality of the effluent. C. Failure to Report Bypass Incidents 109. At no time during this seven-week period of bypass incidents did defendant notify (either orally or in writing) DOH, the public or any other entity as required by its permit and the mandatory procedures binding upon it. Thus, DOH was not given the opportunity to implement measures to mitigate the bypasses, to notify the public, or to eliminate the public health dangers, if any existed. T10. Plant employees have testified that there were several reasons for their failure to report the bypasses to the DOH. First, they believed that they were not required to report the bypasses because the interim effluent limitations had not been exceeded. Moreover, Norman Ching, the plant superintendent at the time of the bypasses, testified that he considered the maintenance done on clarifier #1 to be essential. In addition, Ching also testified that he believed that the city had adequately notified the DOH by its letter dated July 27, 1989, which warned of possible hydraulic overloading and reduced efficiency. 111.Management level plant employees, including the plant superintendent, either were unaware of, or disregarded, the bypass reporting provisions of the permit. They did not believe that they were required to report the bypasses to DOH. Lower level plant employees, some of whom expressed concern over the bypasses to their supervisors, were told that the plant had “permission” to bypass sewage (by wastewater processing superintendent Allen Perry (“Perry”) or assistant plant superintendent Nicanor Músico (“Músico”)). They were also told by their supervisors that no reporting of the bypasses was necessary. Nonetheless, lower level plant employees did express concern about the appropriateness and legality of the bypasses. Plant employees never received any training, or written, or verbal instruction on how to minimize or report bypasses. 112. Ching was on vacation when the bypasses commenced on October 3,1989; assistant plant superintendent Músico was in charge of the plant at that time. Ching returned to work sometime in late October 1989. 113. Plant employees testified that they believed that it was impossible to have bypasses at the Honouliuli plant. Superintendent Ching, for example, testified that in October and November 1989, he understood that a bypass involved the diversion of in-fluent sewage around the entire processing plant to be discharged directly through the outfall. He testified that he believed it was “impossible” to have bypasses at the plant because the plant did not have a bypass pipe. Similarly, Músico testified that he believed that what happened at the bypass weir was an overfloiv, which apparently involves the spilling over of sewage from a tank or channel and was not required to be reported as a bypass to the DOH. 114. As discussed above, the city’s internal reporting procedures require the immediate reporting of any “emergency or anticipated discharge of raw or partially treated sewage (overflows/bypasses).” Thus, the distinction between bypasses and overflows was irrelevant for reporting purposes, at least within the WWT & D. Since the purpose of the internal reporting requirements was to ensure adequate notification to the DOH and the public, it is logical to conclude that the city intended the plant to report any overflow or bypass, and that the appropriate city offidal would then inform the DOH and the public. The notification letter that the city sent to the DOH regarding the maintenance of primary clarifier # 1 does not mention the possibility of either bypasses or overflows. Yet, as Músico testified, the city was aware that overflows might occur and that the interim effluent limitations might be exceeded. 115. The trial testimony of the plant supervisors and employees and the evidence before the court demonstrates that these individuals were ivell aware of the term “bypass” and of its meaning. Plant employees were familiar with, and regularly used, the term “bypass” as a reference to what Músico called an overflow. In the internal wastewater bypass report submitted by Ching on July 19, 1989, for example, he referred to “possible in-plant bypass of partially treated wastewater.” 116. Other city employees in the Waste-water Management division (“WWMD”) were well aware that the NPDES permit defined a bypass as an intentional diversion around any portion of the treatment process. Moreover, they were not aware that the plant employees believed that the flows over the bypass weir were overflows, not bypasses. 117. The city’s Water Quality section, which is part of the planning branch of the WWMD, acted as the “keepers” of the Honouliuli NPDES permit. The Water Quality section is responsible for administering the city’s NPDES permits. David Nagamine, an engineer in the Water Quality section, testified that the practice of his section is to give copies of the city’s NPDES permits to the Wastewater Treatment and Disposal branch (“WWT & D”), which is responsible for administering the treatment plants. The WWT & D was responsible for informing the individual plants of the requirements under their NPDES permits. There was some testimony that the main plant at Sand Island may have received copies of the NPDES permit for the plants in the area. 118. The city failed to keep a copy of the Honouliuli NPDES permit at the plant itself. Ching and Músico were only vaguely familiar with the bypass provisions in the permit and never asked to see the permit. They were apparently operating under the mistaken assumption that the terms of the 1985 consent order had superseded all the provisions contained in the NPDES permit. In this case, the city never sufficiently coordinated the keepers of the permit with the operators of the Honouliuli plant. D. DOH Investigation 119. No city employee ever reported the bypasses to DOH or EPA. In fact, these unreported bypass incidents were not discovered by DOH, EPA, or the public until an anonymous tip by a plant employee on or about November 17, 1989, to a member of plaintiff Sierra Club, who in turn immediately reported the incidents to DOH enforcement officer Ciesla. Ciesla, a DOH (and EPA) employee with extensive experience monitoring compliance of sewage plants, was responsible for investigating noncompliance allegations and for making recommendations to DOH about appropriate enforcement action. 120. City employees cooperated in the DOH investigation of the bypass incidents by speaking with Ciesla and preparing a chart documenting the amount of sewage bypassed per day. 121. DOH later initiated a limited administrative enforcement action against the city for failure to sample effluent on 22 weekend days during the October-November 1989 period, since the city could not prove that the interim effluent limitations were met on those days. DOH issued a Notice and Finding of Violation and levied $449,000 in fines against the city, However, the city has yet to pay any fines to DOH as a result of this administrative action. E. Bypass and Reporting Violations 122. During these bypass incidents (and at all times from July 1, 1988 to present), the effluent limitations in the plant’s permit are pegged to secondary treatment of sewage (including the 85% removal rate). 123. In its July 3, 1991 order, the court held that the city violated the permit and the Clean Water Act on 52 separate dates by bypassing sewage in contravention of the permit provision. Because the city failed to meet the required secondary treatment levels during the bypasses, the bypasses were prohibited regardless of whether the repair of clarifier # 1 was essential. Even if the city had been bound only by the interim effluent limitations in the consent order, it would still be hable for the bypasses because it failed to establish that the maintenance was essential. There appears to be no reason that it was necessary to begin the maintenance in October, when flows are predictably seasonally high. 124. The court further held in its July 3, 1991 order that the city violated the permit on the same 52 separate dates by failing to report immediately the bypasses to DOH, the public or other required governmental entities, and by failing to file the mandatory written reports within five days thereafter. The fact that plant employees may have believed they were not required to report the bypasses because the discharges remained within the interim effluent limitations does not relieve the city of its statutory obligation, although it may be relevant to determining whether the city acted in good faith. 125. In addition, the city had anticipated possible bypasses during the repairs of primary clarifier # 1, but failed to adequately notify DOH in advance of the bypass. This failure to give prior notice also violated the terms of the NPDES permit. 126. Both the city’s and plaintiffs’ experts agree that, if the city had in place during the bypass episodes the recently completed 13 mgd plant expansion, the bypasses would not have occurred at the flow levels then experienced. In addition, the city has made several improvements in the bypass weir itself, which, if they had been done before October 1989, could have prevented the bypasses. In January 1990, the city installed a 24 inch high marine wood weir between the influent and effluent channels. (The bypass weir in 1989 was eight inches high.) This temporary measure was approved by DOH as able to withstand peak flows of up to 40 mgd for short periods of time. Then in January 1991, a concrete wall was built to replace the wood weir. 127. No bypasses have occurred at the Honouliuli plant since November 1989. 128. Plaintiffs have demonstrated that, at the time their complaint was filed, there was a likelihood of recurrent bypass violations at the plant and, accordingly, plaintiffs have satisfied the Gwaltney doctrine. III. SECONDARY TREATMENT VIOLATIONS 129. As discussed above in Section I.G.2, the city has operated the Honouliuli plant as if the interim effluent limitations established in the 1985 consent order have been in effect even after July 1, 1988. The court has ruled, however, that neither the EPA nor the DOH has the authority to extend the compliance deadline for secondary treatment beyond the statutory deadline of July 1, 1988. July 3, 1991 Order at 21-23. The court also held that the pendency of a 301(h) waiver application does not shield the defendants from liability for violations of the secondary treatment requirement. Id. at 22. Accordingly, by order dated May 8,1992, the court granted summary judgment in favor of plaintiffs, finding that the city had been in continuous daily violation of the secondary treatment effluent limitations. 130. The Honouliuli plant has exceeded all of the secondary treatment effluent limitations on the discharge of BOD and SS contained in its permit every day from July 1, 1988 to December 31, 1992. These limitations are: the 30 day average and the 7 day average for the mass emission rate and concentration, and the percent removal. IV. FACTORS RELATED TO ASSESSMENT OF PENALTIES A. Seriousness of Violations 131. In evaluating the seriousness of the city’s bypass and secondary treatment violations, the court looks to several factors, including, but not limited to: (1) the number of violations; (2) the duration of noncompliance; (3) the significance of the violation (degree of exceedance and relative importance of the provision violated); and (4) the actual or potential harm to human health and the environment. (See EPA, “Clean Water Act Penalty Policy,” Feb. 11, 1985, at 3-5.) 1. Number of Violations 132. The number of violations of the Clean Water Act in these cases is significantly high. The city has committed 104 bypass and reporting violations, and daily secondary treatment violations on 1645 days from July 1, 1988 to December 31, 1992. 2. Duration 133. Both the bypass and secondary treatment violations are of long duration. The bypass and reporting violations continued on a daily basis for six weeks in October and November 1989. Moreover, there has been four-and-a-half years of continuous secondary treatment violations, commencing on July 1, 1988 and continuing to the present day. 3. Significance 134. The bypass and reporting violations are significant violations both in terms of the degree of exceedance of the permit limitations and the importance of the permit provision violated. 135. The mass emission rate limitation for BOD and SS is 2840 kg/day (monthly average). Comparing these permit limitations to the actual emissions during October 3 through November 28,1989, it is evident to the court that the bypassed effluent made a significant contribution to the plant’s overall exceedance of its permit limitations. 136. Ciesla testified that he could not recall any series of bypass episodes in Hawaii that ever exceeded the 106 million gallons bypassed in this case. 137. The prohibition against bypass is an important provision of the NPDES permit because a bypass involves the discharge of sewage that has skipped a unit of treatment. Fortunately for the city, the effluent never exceeded the interim limitations established in the 1985 consent order. Since the interim limitations do not apply after July 1, 1988, however, the city has been in continuing violation of its NPDES permit effluent limitations. 138. The city did not follow the detailed provisions of its own directive and standard operating procedures. Although an appropriate internal wastewater bypass report was submitted prior to the anticipated bypass incidents, the city failed to inform DOH either before or during the incidents. 139. The failure to report is even more significant than the bypasses themselves because without notice, DOH was deprived of the critical opportunity to investigate in a timely manner, direct the city to take mitigative measures, warn the public through the media, post signs, or require additional monitoring. The city had a duty to inform DOH of the bypasses even if it believed in good faith that no harm to public health or the environment would have resulted. Testimony revealed that jolant managers felt that DOH had no role to play in the internal operations of the plant. The court finds this attitude troubling, to say the least, because neither the city nor the plant managers are in sole control of the operation of the plant. The reporting requirements are designed to afford DOH an opportunity to provide a “check” on the operations of the plant. Even if DOH would have determined that no protective measures were needed, it still should have been informed. 140. Similarly, but to a much larger degree, the city’s secondary treatment violations are significant in terms of the degree of exceedance of the permit limitations and the importance of the provision violated. 141. The city’s daily secondary treatment violations resulted in discharges significantly above the NPDES permit levels for BOD and SS. 142. The secondary treatment-based effluent limitations in the city’s permit are a fundamental requirement of the permit. That the city believed the interim limitations had superseded the NPDES permit limitations does not detract from the significance of the violation, although it may be relevant to whether the city acted in good faith. 4.Harm to Environment and Public Health 143. The Honouliuli effluent consists of 23-25 mgd of primary-level treated sewage. In 1992, the Honouliuli plant discharged approximately 11,000 kg (or 24,200 lbs) of BOD every day and 4,500 kg (9,900 lbs.) of SS. The NPDES permit limit is 2,840 kg/day of BOD and 2,840 kg/day of SS. 144. The plant currently removes approximately 40-50% of the BOD and 70-80% of the SS from the influent; it is operating effectively as a primary treatment system. 145. If the plant provided secondary level treatment, the removal rates would probably be approximately 90-95% for both BOD and SS. 146. Sewage consists of both organic and inorganic compounds and particles of various sizes. Sewage consists of sediment, nutrients, (e.g., nitrogen, phosphorous, nitrites and nitrates), freshwater, and organic materials known as “biochemical oxygen demand” (which reduce the oxygen available to the ambient ocean environment). 147. The effluent of the Honouliuli plant consists primarily of domestic sewage, which can contain a variety of pathogenic organisms, including both bacteria and viruses. Sewage particles (both SS and organic materials) can provide a transport vehicle for pathogens. 148. The Honouliuli plant does not disinfect, and has never disinfected, its effluent. If the plant were to disinfected its effluent, a high percentage of the pathogens could be killed; however, the plant is not currently required by its permit or by the DOH to disinfect its effluent. The plant is equipped with chlorination equipment, but the DOH has never required the plant to disinfect its effluent because the state water quality standards have not been violated in the receiving waters. Of note, however, is the fact that the plant is not presently equipped with dechlorination facilities to remove chlorine from the disinfected effluent before it is discharged, in the event chlorination were required or ordered. 149. The city currently conducts comprehensive monitoring of the water quality at the plant and in the receiving waters, of the biological community in the area of the outfall, of the sediment in the vicinity of the sewage discharge, and of the outfall/diffuser itself. Relevant data from this monitoring program is discussed below in connection with the impact of the Honouliuli discharge. a. Fate of the effluent seivage 150. Plaintiffs have presented testimony that the Honouliuli outfall and diffuser system is currently operating outside of its design parameters. Dr. Scott Jenkins (“Jenkins”), plaintiffs’ expert in physical oceanography, specializing in coastal processes and hydraulics, has criticized the plant as being designed for secondarily treated effluent at higher flows. Because the plant discharges primarily treated sewage at lower rates of flow, he concluded that there must not be sufficient pressure to diffuse the sewage to achieve the dilution of the effluent predicted by the city’s experts. 151. Although the plant may not be operating at optimal design levels, there is no evidence before the court that it is not operating efficiently and effectively. The design range of the outfall is not as narrow as Jenkins posits. The Barbers Point outfall design report states, for example, that “[t]he optimal outfall design for the Barbers Point Outfall System is practically identical whether considering primary or secondary effluent.” And the design report goes on to explain that an “additional 600 feet of diffuser has been included for the flexibility of using the ocean outfall for either secondary or primary effluent.” The court is not in the business of second guessing the proper pipe size for a deep ocean outfall. The court recognizes that the construction of the outfall was constrained by costs and feasibility as well as optimal design features, and does not find that the current outfall was the result of an ill-informed decision by the city. The court finds that the evidence shows that the removal rates .for both SS and BOD are above the 30% rate that defines primary treatment. Moreover, surveillance of the outfall and diffuser pipes shows no malfunction in the diffuser ports or other operational problems. 152. The sewage discharged through the outfall contains fresh water, which is lighter than sea water; therefore, the sewage plume tends to rise as it leaves the outfall. The sewage plume usually remains submerged and spreads horizontally when it reaches water that is equal to it in density. At other times, however, the plume enters the top mixed layer and surfaces. Dr. Edward Noda (“Noda”), the city’s expert in coastal engineering and hydrodynamics, has estimated from available data that the sewage plume from the Honouliuli outfall surfaces approximately 25% of the time over a year. There are, however, well-established seasonal variations in the depth of the mixed layer, so that surfacing occurs about 15-25% in the summer, 30-40% in the spring, and 50% or more in the winter (it has gone as high as 98% in November). 153. Noda presented a computer generated model illustrating the fate of the sewage plume as it leaves the diffuser. The Noda model is based on water density, wind and current data taken near the Honouliuli outfall. Based on this data, Noda calculated the initial dilution, in the worst case scenario, to be 118:1 for a submerged plume and 461:1 for a surfacing plume. These figures may be not reflect the fact that episodes of less dilution could occur, because the Noda model predicts averages over time rather than specific instances. The model does give an overall picture of the fate of the sewage plume, however. 154. The Noda model predicts that effluent from the Honouliuli outfall will surface and reach the shoreline along Mamala Bay approximately .1% of the time over a year. The effluent that reached the shore would be diluted at minimum approximately 1700:1. Not surprisingly, plaintiffs’ experts suggest that Noda’s model underestimates the frequency of surfacing and the extent of shoreward transport. Jenkins and Dr. Douglas Segar, plaintiffs’ oceanography expert specializing in the cross disciplinary investigation of marine pollution, offered credible testimony about wind and current effects that suggests the net direction of transport during both Kona and tradewind conditions is northwest toward shore rather than off to the southwest as Noda had predicted. Dr. Segar also testified that actual current data shows that the current speed was slower than Noda assumed, which would lead to less dilution when the plume reached the shoreline. Their testimony is supported by current measurements taken near the Honouliuli outfall, as reported in the city’s 1983 reapplication for a 301(h) waiver. 155. Accepting that the Noda model may underestimate the frequency of the plume reaching shore, and overestimate dilution, there is still no evidence that counters Noda’s conclusion that there has been no actual harm to the environment or to public health. Plaintiffs have made an arguable case for