See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/233656616 Electrolytic Sodium Hypochlorite System for Treatment of Ballast Water Article in Journal of Ship Production · August 2006 DOI: 10.5957/jsp.2006.22.3.160 CITATIONS READS 32 6,869 8 authors, including: Jeffery Cordell Bryan Nielsen University of Washington Seattle WaterTectonics 139 PUBLICATIONS 3,014 CITATIONS 4 PUBLICATIONS 36 CITATIONS SEE PROFILE SEE PROFILE Nissa C Ferm David J. Lawrence National Oceanic and Atmospheric Administration National Park Service 5 PUBLICATIONS 202 CITATIONS 33 PUBLICATIONS 1,248 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Seattle Seawall Project View project Onsite sodium hypochlorite generaton for treatment of ballast water View project All content following this page was uploaded by Jeffery Cordell on 01 October 2015. The user has requested enhancement of the downloaded file. SEE PROFILE Journal of Ship Production, Vol. 22, No. 3, August 2006, pp. 160–171 Electrolytic Sodium Hypochlorite System for Treatment of Ballast Water Rudolf C. Matousek,* David W. Hill,* Russell P. Herwig,† Jeffery R. Cordell,† Bryan C. Nielsen,† Nissa C. Ferm,† David J. Lawrence,† and Jake C. Perrins† *Severn Trent DeNora, Sugar Land, Texas, USA † University of Washington, Seattle, Washington, USA The potential problems of organisms introduced by ballast water are well documented. In other settings, electrolytic generation of sodium hypochlorite from seawater has proven to be a simple and safe method of handling and injecting a biocide into water. After the hypochlorite oxidizes organisms, it reverts back to the chloride ion. Mesocosm-scale testing of this technology combined with filtration, using organisms from Puget Sound, Washington, demonstrated that hypochlorite generation and use may be a viable method to eliminate aquatic nuisance species from ballast water while minimizing disinfection byproducts and residual toxicity. These experiments were conducted at the U.S. Geological Survey Marine Field Station on Marrowstone Island, Washington. Results from the first set of studies of the system showed that hypochlorite levels greater than 3.0 ppm hypochlorite with or without filtration reduced bacteria by more than 99.999%, reduced phytoplankton by more than 99%, and reduced mesozooplankton by more than 99%. Filtration improved efficacy only when hypochlorite concentration was initially less than 1.5 ppm. 1. Introduction 1.1 Regulations THE MARINE ENVIRONMENT PROTECTION COMMITTEE (MEPC) of the International Maritime Organization met in February 2004. The committee adopted a new world Ballast Water Convention that will enter into force 12 months after ratification by 30 member states representing 35% of the world’s gross tonnage. The Convention is divided into articles and, more importantly for the development of ballast water treatment technologies, an annex that includes technical standards and requirements in the regulations for the control and management of ship ballast. All ships, including submersibles, floating craft, floating storage units (FSUs), and floating production, storage, and offloading units (FPSOs), are to manage their ballast water in accordance with an approved ballast water management plan and record such management in a ballast water record book. All ships greater than or equal to 400 gt are to be surveyed (initial, annual intermediate, and renewal) and Presented at the 2005 Ship Production Symposium, Society of Naval Architects and Marine Engineers, October 19 to 21, Houston, Texas. 160 AUGUST 2006 certificated (not exceeding 5 years). Table 1 summarizes the implementation schedule of the type of treatment required according to the age of ship and its ballast capacity as per the provisions of the Convention (International Maritime Organization 2004). Indicator microbe concentrations shall not exceed: • Toxicogenic Vibrio cholerae: 1 colony forming unit (cfu) per 100 ml • Escherichia coli: 250 cfu per 100 ml • Intestinal Enterococci: 100 cfu per 100 ml. Ballast water exchange is to take place as follows: • At least 200 nautical miles from the nearest land and 200 m water depth OR • In the event throughout the intended route the sea area does not afford the above characteristics, in a sea area designated by the port state. There may be a need to alter the ship’s intended route to exchange ballast in the designated area. States may establish additional ballast water management measures for ships to meet based on guidelines, which remain to be 8756/1417/06/2203-0160$00.49/0 JOURNAL OF SHIP PRODUCTION Table 1 Ballast Capacity (m3) <1,500 >1,500 <5,000 >5,000 Construction Date Ballast water treatment implementation schedule 2009 2010 <2009 >2009 <2009 >2009 <2012 >2012 2011 2012 2013 2014 2015 2016 D1 or D2 D2 2017 D2 D1 or D2 D2 D2 D1 or D2 D2 D1 ⳱ ballast water exchange (95% volumetric exchange) or pumping through three times the volume of each tank. D2 ⳱ ballast water treatment systems approved by the Administration with treatment efficacy of: + Not more than 10 viable organisms per m3 >50 m. + Not more than 10 viable organisms per milliliter >10 and <50 m. developed. The MEPC shall undertake a review of the Ballast Water Standards no later than 2006 and is to include an assessment of the technologies available that achieve the standard. As part of the assessment, the MEPC requires significant documentation of the system, with efficacy data for both land-based and shipboard tests. Criteria for such testing have been established and outlined in MEPC document MEPC 52/WP.7 titled “Guidelines for Approval of Ballast Water Management Systems.” As a result of these regulations, there is a need for a proven, viable, cost-effective ballast water treatment system by 2007. This will allow ship designers to specify and incorporate such devices into ships that begin construction after January 1, 2009. This paper is the summary of land-based work on a ballast water treatment (BWT) system utilizing on-site hypochlorite generation. 1.2. Ballast water management strategies 1.2.1. Open-ocean ballast water exchange. Most environmental scientists agree that ballast water that is exchanged in the open seas presents less of an ecological threat or risk to the receiving waters. The question is whether the threat has been reduced to an acceptable level. As presently engineered and practiced, reballasting does not exchange 100% of the ballast water or remove all of the sediments that are found in ballast water tanks. The amount of the exchange varies from ship to ship. Older ships tend to exchange less efficiently than newer ships. Dickman and Zhang (1999) examined four containerships that took on ballast in Mexico and discharged 21 days later in Hong Kong. After this period, few of the dinoflagellate and diatom species taken on in Mexico were alive in the ballast water in Hong Kong. Five ships that reballasted in the open ocean reduced the diatom and dinoflagellate populations by 48%. They concluded from results of this and previous study that older vessels were less effective in removing diatom and dinoflagellate species than newer ships. The reason could be that the reballasting design of older ships was less efficient in removing water and sediments located near the bottom of the ballast tanks and that the bottom water is associated with a large number of resting cysts and cells. Ships sampled in Puget Sound by the University of Washington’s ballast water team showed that even in ships that reported open ocean exchange, up to 50% of the organisms were nonnative to the northeastern Pacific. 1.2.2. Ballast water treatment technologies. A variety of treatment technologies are suggested for the removal or reduction of AUGUST 2006 organisms that are found in ballast water. The challenge is to develop a technology that is effective against a variety of biological taxa and that is capable of quickly treating the very large volumes of ballast water that are associated with large ships. Treatment technologies may be developed for on-board or at-port applications. Technologies that have received most of the recent consideration include (1) cyclonic separation, (2) heat treatment, (3) electric pulse, (4) ultraviolet light, (5) filtration, and (6) biocides. These technologies were listed and reviewed in the National Research Council publication “Stemming the Tide: Controlling Introductions of Nonindigenous Species by Ships’ Ballast Water” (National Research Council 1996). We briefly review the proposed technologies and provide some comment. Cyclonic separation. A more sophisticated and more recently developed technology involves cyclonic separation. This is normally accomplished using hydrocyclones. If properly designed and applied, hydrocyclones will require less pump pressure than screen filters and will allow separation of sediments and other suspended solids to approximately 20 µm. Hydrocyclones are limited, however, to separating solids with a specific gravity greater than, or less than, water. Many types of organisms (e.g., bacteria and other microorganisms) will not be separated from the water since their specific gravity is extremely close to seawater. Ultraviolet light. In close contact, ultraviolet (UV) light treatment is effective in killing non–spore-forming microorganisms. It does not appear effective in inactivating higher organisms or the cyst or resting stages of protozoa. Ultraviolet light was not effective in totally eliminating dinoflagellates. Montani et al. (1995) found that after exposure to ultraviolet light for 2 hours, the germination of Chattonella sp. cysts decreased to 6% of the control, whereas germination of cysts of other species of dinoflagellates (Alexandrium sp. and Gymnodinium sp.) was more than 40% of the controls. The University of Washington research team has performed controlled mesocosm tests at the Marrowstone Marine Field Station in which Puget Sound seawater was amended with additional zooplankton. The amended seawater was exposed to moderate levels of UV light. They found that zooplankton mortality was delayed and did not immediately occur. The microbial populations were quickly reduced in number following exposure to the UV light but rebounded to their original levels within a few days. Heat treatment. The use of waste heat from the ship’s propulsion and service cooling is an attractive option for the inactivation of organisms in ballast water. No chemical byproducts or residuals JOURNAL OF SHIP PRODUCTION 161 would be associated with the discharge of heat-treated ballast water. A number of factors will limit the practicality of thermal treatment primarily related to the volume of water that is associated with large vessels and the amount of energy required to heat the volume. Thermal treatment may be more applicable to ballast water originating from warmer environments. Heat required for thermal treatment could be reduced where water temperatures are at tropical or summer levels (30 deg C or higher). The heat loss to the ambient waters outside of the hull must be considered. Different types of organisms or organisms from different parts of the world may have different sensitivities to heat. A few recent studies suggest that heating ballast water is the best method for killing a variety of higher organisms and microorganisms that are found in ballast water. Rigby et al. (1999) showed how a cost-effective heating technique using waste heat from a ship engine could be used to kill many unwanted organisms. In an ocean trial, heated water flushed through one of the ballast tanks resulted in the destruction of all zooplankton with very limited survival of the original phytoplankton. Electric pulse. Small-scale experiments have been performed by applying electrical voltages in the 15 to 45 kV range with pulse duration of 1 µs. Large energy sources would be required for systems capable of treating large volumes of ballast water (National Research Council 1996). Filtration. The physical separation and removal of organisms can be accomplished during ballasting operations using a shipboard filtration system. Some would argue that this technology is the most promising choice. Filtering ballast water as it is loaded is an attractive option since it would minimize the introduction of unwanted organisms. The options for onboard filtration systems are either mesh strainers or deep media filtration. Many problems associated with a strainer technology have been solved with the development of commercially available continuously cleaned screening systems. Media filters are attractive in principal because small-size particles can be removed, but these filters are most likely unrealistic for on-board treatment of ballast water because of the large footprint that would be required. The primary disadvantage of the strainer filters is that many organisms are smaller than strainers and would pass through the treatment system. A study for the Canadian Coast Guard (1992) concluded that the physical removal of organisms by filtering may be an effective stand-alone treatment process or may be used in conjunction with other technologies, such as chemical treatment or UV sterilization. Flow-through centrifugation systems can separate particles prior to filtering to reduce filter clogging. Biocides. The addition of chemicals that would kill or inactivate a variety of organisms found in ballast water is an attractive treatment technology because of the ease of application. A biocide could simply be added to the ballast tank and allowed to react for a specified period. Biocides are among the most widely used industrial chemicals, and there is a large body of knowledge about their use in wastewater treatment. If similar concentrations were required to inactivate organisms found in ballast water, then a large ship would need to carry only a few cubic meters of biocide per voyage. The use of biocides for ballast water treatment has been rejected by some for several reasons, including the reluctance to add toxic chemicals to water that may be discharged back into the ocean, the unknown effectiveness of biocides against target organisms, and compliance with discharge regulations around the world. Oxidizing biocides such as chlorine, chlorine salts, and 162 AUGUST 2006 ozone have been used for decades in a variety of sanitizing applications. Ozone is a strong oxidizing agent used for treatment of potable and industrial waters. With the increasing environmental concern associated with the use of chlorine, ozone has received greater attention in recent years. Ozone is an unstable gas that must be generated as needed, and some reviewers have concluded that ozone may not be practical for shipboard use (National Research Council 1996). Along with generation, ozone efficacy is a problem due to the gas–liquid contact requiring elaborate diffusion equipment. In salt water, ozone produces many of the similar residual compounds as chlorination. Nonoxidizing biocides such as glutaraldehyde or vitamin K have also have been suggested. 2. Electrochlorination 2.1. Background Seawater (normally between 15 and 35 grams/liter) or other water containing NaCl may be used to generate a disinfecting solution containing chlorine by passing a direct electrical current through the solution. On-site generation of hypochlorite from seawater has been used for over 25 years. These systems can be purchased as completely skid-mounted systems that generate sodium hypochlorite from seawater. These systems are used in refining, petrochemical power plants, offshore drilling production, and marine applications around the world. Systems can be scaled to the appropriate size depending on the quantity of hypochlorite required. The type of electrolytic cell commonly used in these marine and offshore applications is a “tube within a tube.” A cell consists of one anode, one cathode, and one bipolar tube with the necessary ancillary hardware to facilitate assembly. The outer anode and cathode are manufactured from seamless titanium pipe. The anode surface is coated with proprietary precious metal oxides, primarily ruthenium and iridium. Seawater enters one end of the cell and passes between the cathode, the anode, and bipolar tube annular spaces. When direct current is applied to the cell, sodium hypochlorite results. One cell can produce up to 5.5 kg/day, and a maximum of 12 cells can be connected in series for a capacity of 65 kg/day per train. In some applications, such as ballast water treatment, a dechlorination step can be added to the process. This requires adding a reducing agent, such as sodium sulfite, to the end of the system to neutralize any residual chlorine at the point of discharge. The end result is a nontoxic stream with no free chlorine. 2.2. Chemistry The process is based on the partial electrolysis of NaCl present in seawater as it flows through an unseparated electrolytic cell. The resulting solution exiting the cell is a mixture of seawater, sodium hypochlorite (hypo), hydrogen gas, and hypochlorous acid. Electrolysis of sodium chloride solution (seawater in this study) is the passage of direct current between an anode (positive pole) and a cathode (negative pole) to separate salt and water into their basic elements. Chlorine generated at the anode immediately goes through chemical reactions to form sodium hypochlorite and hypochlorous acid. Reactions are shown below: Cl− → Cl2 (aq) + 2e− Eo ⳱ 1.396 V (1) JOURNAL OF SHIP PRODUCTION which is hydrolyzed in solution to form hypochlorous acid: Cl2 + 2H2O → 2HOCl + 2H+ (2) Hypochlorous acid dissociates to hypochlorite at alkaline pH levels: HOCl → OCl− + H+ pKa = 7.5 (3) In seawater bromide ions are present, together with a range of inorganic cations as well as possibly ammonia and a variety of organic compounds. The reaction of molecular chlorine or hypochlorite ions with ammonia or amino compounds leads to the disinfectants, and they react to destroy bacteria and microorganisms in the water just as do chlorine, hypochlorous acid, and hypochlorite ions. The rapid oxidation of bromide ions will also occur and (as with chlorine) in the aqueous environment form hypobromous acid (HOBr) and hypobromite ions (OBr−). These reactions will also be equilibrium processes, dependent on temperature and pH. Note also that brominated species will react with ammonia and/or amino compounds if present in the water, just like the chlorine analogues. Hydrogen and hydroxides are formed at the cathode, the hydrogen forms a gas and is vented, and the hydroxide aids in the formation of sodium hypobromite and increases the exit stream pH to approximately 8.5. This reaction is shown as follows: 2H2O + 2e− → H2 (g) + 2OH− Eo = −0.828 V (4) Because the electrolytic cell used for this application is unseparated, the reactants at both anode and cathode can further react to form the respective end products shown in the overall electrochemical and chemical reaction as follows: NaCl + H+ + Br− + 2e → NaOBr + H2 + Cl− Salt + Water + Energy → Sodium Hypochlorite + Hydrogen (5) 2.3. Disinfection by-products Disinfecting agents, such as chlorine, ozone, chlorine dioxide, and chloramines, react with natural organic material present in water to produce disinfection by-products (DBPs). Most of the research and interest in DBPs has been with drinking water. DBPs have been known since 1974, when chloroform was identified as DBP resulting from the chlorination of tap water. Since then, hundreds of DBPs have been identified in drinking water. The benefit of disinfecting drinking water is obvious, as thousands of people died from waterborne disease before municipalities began to disinfect drinking water, but it is also generally recognized that it is important to minimize the formation of DBPs in drinking water. Several DBPs have been linked to cancer in laboratory animals, and as a result, the US Environmental Protection Agency (EPA) has regulated some DBPs. While we anticipate that most people will not be drinking ballast water and other treated seawater, we have evaluated the formation of selected DBPs that may be formed following the generation of hypochlorite in seawater. Seawater is significantly different from freshwater, not just because of the relatively high concentration of Na+ and Cl- and oftentimes higher levels of natural organic material, but also because of the presence of Br− (bromide). The presence of this ion may lead to the formation of bromate (BrO3−), a compound that is considered a possible human AUGUST 2006 carcinogen. In the United States, bromate is regulated at 10 µg/L (10 parts per billion) in drinking water. Seawater contains a typical bromide concentration of 65 mg/L, so the concentration in seawater is significant. Therefore, bromate was one of the DBPs measured in our research. In addition, the presence of haloacetic acids (HAAs) and trihalomethanes (THMs) are a concern and were measured as part of the overall study. One other component of the ballast water treatment system is the neutralization of the free halogen (hypochlorite and hypobromite) prior to discharge from the ballast tanks. Sodium sulfite is used, and the simplified reaction is shown below to form sodium sulfate. As shown in the equation, neutralization occurs at one to one molar ratio but two to one (sulfite to halogen) as a weight ratio. Na2SO3 + Cl2 + H2O –> Na2SO4 + 2HCl Sodium Sulfite + Chlorine + Water –> Sodium Sulfate + Hydrochloric Acid (6) Typically, there are 4 g/L sulfate in seawater, and based on the concentrations of halogen required, only 10 mg of sulfate will be added to the discharge ballast water. Also, the amount of HCl generated is negligible and will not change the pH of the discharge ballast water. 3. Pilot electrochlorination ballast water treatment 3.1. Background The University of Washington School of Aquatic and Fishery Sciences performed third-party verification tests on the Severn Trent DeNora’s Electrochlorination Ballast Water Treatment (BWT) System (BalPure; Severn Trent DeNora, Sugar Land, TX). The pilot plant system consisted of two 5.7 m3 (1,500 gallon) raw seawater holding tanks. The water was pumped through a 50-m self-cleaning filter. When the pressure drop reached a preset value, the filter was automatically back flushed while continuing to operate. A stream volume of roughly 10% of the inlet flow to the filter was generated, containing the removed organisms and solid contaminants. During a ballast water operation, this stream would be discharged overboard. For purposes of this test, the stream was collected. After the water was filtered, a side stream was fed to the electrolytic cell, where oxidizers were generated. This oxidized stream was then injected and mixed with the main stream. After the hypochlorite stream was mixed with the mainstream, the water was sampled and the free halogens were measured and recorded. This on-line real-time free halogen value was used to automatically adjust the amount of hypochlorite generated for the ballast water treatment. Based on previous lab tests, the expected required dosage after filtration ranged from 1 ppm up to 5 ppm. Once the water was treated, it was placed in replicate dark mesocosms. This step was analogous to ballast water in ballast tank conditions. After 7 to 14 days of storage, the seawater was monitored before it was “discharged.” A dehalogenation agent, such as sodium sulfite, was injected into the water to react or neutralize any free halogens. A schematic of the pilot process is shown in Fig. 1. JOURNAL OF SHIP PRODUCTION 163 Fig. 1 Pilot electrochlorination treatment system 3.2. University of Washington test program 3.2.1. Test site. During the past 4 years, the University of Washington has been conducting ballast water treatment technology testing at the USGS Marrowstone Island Marine Field Station. To date, experiments have been conducted with Puget Sound seawater amended with locally obtained zooplankton greater than 73 m. The Marrowstone Marine Field Station is located on Marrowstone Island at the northwestern entrance to Puget Sound, Washington, USA. The station is a former US Coast Guard lighthouse, 164 AUGUST 2006 acquired by USGS in 1974. Besides the old lighthouse keeper’s residence, the station has a laboratory/office building, two wet labs with constant seawater flow, several large tanks, many smaller tanks, a pump house, and shop. Approximately $4 million was invested in construction during the 1990s. Seawater for the laboratory is drawn directly from northern Puget Sound. Seawater effluent from the laboratories can be chlorinated and released into a lagoon, and the station has the required permits for releasing treated seawater. Presently, seawater can be pumped at a maximum rate of 1,500 liters (400 gallons) per minute. The station’s filtration system was bypassed for obtaining untreated seawater, JOURNAL OF SHIP PRODUCTION and large tanks were spiked with additional zooplankton before passage through the electrolytic seawater ballast water treatment equipment. 3.2.2. Research plan. The specific objectives of the research were to evaluate the efficacy of the filtration and hypochlorite treatment system both with and without the filtration system. For the mesocosm experiments (see below), the University of Washington research team conducted experiments with the ballast water treatment system to determine the viability of Puget Sound zooplankton, phytoplankton, and microorganisms and measure the concentration of selected disinfectant by-products after treatment with the BALPURE. 3.2.3. Experimental design and sampling. Four replicate 280 liter (75 gallon) circular tanks were used for each treatment and control condition. The Severn Trent De Nora treatment system incorporates two steps, a filtration step that is followed by the addition of generated hypochlorite. The efficacy of the system was examined separately for each component, and for the filtration and chlorine components used together. For the control, seawater passed through the system without applying the filtration and chlorination steps. Because pumps may destroy some of the zooplankton present in the seawater, separate controls with and without the pump/flow network were used, to determine mortality caused by the pumping system versus that caused by the treatment. To increase the sensitivity of the mesozooplankton analyses, additional mesozooplankton were mixed with ambient Puget Sound seawater for the mesocosm experiments. Mesozooplankton were collected in Kilisut Harbor, Washington, using a 1 m diameter, 110 m mesh net, approximately 2 hours prior to the start of the mesocosm experiment. Mesozooplankton were added to the 1,500 gallon supply tank to obtain a density of approximately 150 to 200 individual organisms per liter. A Hensen-Stempel pipet was used to collect three randomized 5 ml samples of the mesozooplankton concentrate from the collecting vessel after vigorous stirring. The counts from these subsamples were used to calculate how much of the concentrate was needed to achieve 150 to 200 mesozooplankton per liter for the start of the experiment and for restocking mesocosms after the dechlorination step. The appropriate number of zooplankton were mixed with the 1,500 gallon supply tank and allowed to acclimate for at least 1 hour before they were pumped through the treatment system. After thoroughly mixing the contents of mesocosms, plankton samples were collected using 1 L Nalgene wide-mouth bottles. Each sample was filtered through a 73 m sieve and placed in a counting tray. Samples were observed with a stereomicroscope. Each organism observed was classified into one of eight taxonomic groups and categorized as live (vigorous movement or escape response when probed), dead (no movement observed, no response to probing), or moribund (internal and/or external movement observed, but no escape response to probing). In some experiments, after performing the dechlorination step at 24 hours, the contents of the tanks were pumped through a 73 m mesh screen to remove any remaining live and dead mesozooplankton. The filtered mesocosm water was dosed with fresh mesozooplankton concentrate. Hypochlorite was tested at a range of concentrations, a high value near 5 ppm and a low value near 1 ppm. This was done to begin to understand the minimum amount of hypochlorite needed AUGUST 2006 to achieve proposed discharge standards. Each mesocosm was sampled at standard intervals over the 10 day tests. Samples were analyzed for viability of zooplankton, phytoplankton, and bacteria. Mesocosm experiments with the Severn Trent De Nora system were run and are planned during different seasons (i.e., fall, winter, spring, and summer) to capture a wide range of organisms and environmental conditions. We describe here preliminary results from a portion of our mesocosm experiments that are in progress. 3.3. Monitoring methods The change in viable microorganisms and zooplankton before and after passage through the Severn Trent De Nora system were evaluated using several methods. Since the chemical and physical characteristics of the ballast water influence the effectiveness of treatment, chemical and physical parameters were also measured. Table 2 lists the monitoring methods that were used. Each unit of the system was tested separately and in combination (filtration, hypochlorite, sulfite neutralization). 3.3.1. Hypochlorite and by-product chemistries. Treated seawater samples were collected to determine the concentration of the following disinfectant-related compounds: (1) total residual chlorine (or total residual oxidant [TRO]), (2) trihalomethanes (THMs), (3) haloacetic acids (HAAs), (4) bromate. TRO was measured at the site as Cl2 using a Hach Colorimeter (The Hach Company, Loveland, CO). DPD powder pillows (Hach) were used for the analysis. The DBPs analyzed were trihalomethanes (THMs), five haloacetic acids (HAA5), and bromate. Chlorite analysis was not performed in our study. The THMs analyzed, using USEPA Method 524.2, were chloroform, bromodichloromethane, dibromochloromethane, and bromoform. In addition to the specific THMs, the total THM (TTHM) was also determined. The HAA5s analyzed, using US EPA Method 552.2, were monochloroacetic acid, dichloroacetic acid, trichloroacetic acid, bromoacetic acid, and dibromoacetic acid. Bromochloroacetic acid was also determined. Bromate analysis was performed using ion chromatography, using USEPA Method 317. Samples were collected from one representative tank per treatment and control. Samples were collected in glass containers provided by the analytical laboratories and refrigerated until shipment. Samples were packed with ice in coolers Table 2 Methods used to evaluate treated and control seawater used in mesocosm experiments at the USGS Marine Field Station at Marrowstone General chemical and physical parameters pH Temperature Salinity Nutrients Total organic carbon Selected macrofaunal identification and viability Zooplankton Microbiological methods Total culturable bacteria Chlorophyll a Hypochlorite and by-product chemistry Total residual chlorine or total Toxicity assay residual oxidant (TRO) Whole effluent toxicity (WET) Trihalomethanes (THM) bioassay Haloacetic acids Bromate JOURNAL OF SHIP PRODUCTION 165 provided and shipped by overnight express service to arrive at the analytical laboratories after the first samples were collected from the mesocosm experiment within 7 days. To follow US EPA methods, analyses for THMs and HAAs need to be performed within 14 days of collection and for bromates within 28 days. Bromate samples were sent to and analyzed by Nova Chem Laboratories, Inc. (Oxford, Ohio). THM and HAA samples were sent to and analyzed by Edge Analytical, Inc. (Burlington, Washington). Both of these companies are certified contract laboratories. 3.3.2. Culturable bacteria. Water samples collected were enumerated for culturable bacteria. Bacteria were enumerated as colony forming units (CFU) per liter of water. Colonies were cultured on petri dishes containing a growth medium suited for marine heterotrophic bacteria, called marine R2A agar. The inoculated medium was analyzed for colony formation after 5 days of room temperature incubation. 3.3.3. Chlorophyll a. Chlorophyll a was measured as an indicator of phytoplankton biomass. Water was filtered through glass fiber filters with a pore size small enough to retain phytoplankton cells. Chlorophyll a was extracted from the filters using acetone and then analyzed for fluorescence to determine the concentration in g/L (Holm-Hansen & Reimann 1978). Although chlorophyll a analysis provides an assessment of the reduction in phytoplankton concentration after treatment, it does not provide a direct measure of the viability of phytoplankton after treatment. 4. Test results 4.1. Test 1 4.1.1. Conditions. The first experiment was done with and without filtration at an initial chlorine concentration of 3.5 mg/L. Sample sets were taken at 0, 5, 24, 48, 120, 240 hours following treatment and analyzed for total residual oxidant (TRO), culturable heterotrophic bacteria, chlorophyll a, and zooplankton viability. This provided the benchmark for future studies. 4.1.2. Results and discussion. TRO levels declined steadily throughout the experiment (Fig. 2). TRO in the nonfiltered test tanks dropped more than in the filtered test tanks during the first 5 hours. This was a result of organisms being removed by the filtering step and creating a smaller oxidant demand. After the initial drop, the rate of TRO decay over the next 10 days was the same for both test conditions. Bacteria were greatly reduced in both treatments and showed minimal rebound over the 10 day experimental period (Fig. 3). Statistically there was no difference in number of bacteria for filtered or unfiltered conditions. Chlorophyll a is an indicator of phytoplankton biomass. In the treated seawater, chlorophyll a levels were at or below the detection limit beginning with the sample 5 hours after treatment (Fig. 4). Chlorophyll a levels declined over time in the control tanks because the experiments were conducted in the dark. For the mesozooplankton, there was no significant difference between the results for filtered and unfiltered tests (Fig. 5). There were no living organisms in the unfiltered tanks and only one live polychaete larva in the filtered tanks. 4.2. Test 2 Three separate tests of the BALPURE were conducted over a 6-month period in 2004. The test conditions and results are described below. Fig. 2 166 AUGUST 2006 4.2.1. Conditions. The second test was composed of two experiments, consisting of comparing chlorination without filtration at Total residual oxidant (mg Cl2/L) JOURNAL OF SHIP PRODUCTION Fig. 3 Culturable heterotrophic bacteria the relatively low dose rates of 1.0 ppm and 1.6 ppm. Sampling frequency and analysis were the same as in Test 1. 4.2.2. Results and discussion. The TRO level dropped quickly by approximately 0.6 ppm in the first 5 hours and 1.0 ppm in the first 24 hours. Higher residual TRO persisted in both the higher chlorine dose treatment and the filtration treatment (Fig. 6). TRO disappeared in all treatments by 10 days. Bacteria amounts in the first experiment were reduced only when chlorine was added. Because the amount of chlorine added was so small, the bacteria grew back to levels equal to those before treatment. In Test 2, the chlorine levels of 1.5 and 1.0 ppm were insufficient to keep bacteria from growing after approximately 4 Fig. 4 AUGUST 2006 days. But the rate of bacteria growth was inversely proportional to the initial chlorine concentration. This is shown in Figs. 7 and 8. In other words, a minimum level of TRO needs to be maintained to prevent bacteria from multiplying. Chlorophyll a is an indicator of phytoplankton biomass. As observed in Test 1, control samples had approximately 12 g/L of chlorophyll a, the filtered-only sample 7 g/L, and filtration combined with 1.0 ppm chlorine eliminated all chlorophyll a. In experiment 2, even with only 1.0 ppm chlorine and no filtration, chlorophyll a was at the detection limit within 48 hours. In this experiment, densities of approximately 200 individual organisms per liter were used. Fifty m filtration reduced concentrations to approximately 35 live individual organisms per liter Chlorophyll a (µg/L) JOURNAL OF SHIP PRODUCTION 167 Fig. 5 Live mesozooplankton per liter (greater than IMO convention). Addition of 1.0 ppm chlorine reduced the count to less than 10, and 1.0 ppm combined with filtration eliminated all live zooplankton. 4.3. Test 3 4.3.1. Conditions. The primary objective of Test 3 was to measure disinfection by-products and to examine the residual toxicity of neutralized seawater after chlorination. The benchmark analy- Fig. 6 168 AUGUST 2006 ses, done in Test 1 and Test 2, were conducted. A chlorination level of approximately 3.5 ppm was used to maximize disinfection by-products. Filtration was not used in this experiment. One half of the test set (Group B) was neutralized with sodium bisulfite after 24 hours to simulate neutralization of ballast discharge. Excess sodium bisulfite was added at a sulfite-to-TRO ratio of 3:1 (10.5 mg NaHSO3/L) to ensure that no residual sodium hypochlorite or TRO remained in the water. To evaluate the treated seawater for residual toxicity, the neutralized seawater was filtered with a 73 m filter to remove dead organisms and was replenished Total residual oxidant (mg Cl2/L) JOURNAL OF SHIP PRODUCTION Fig. 7 Experiment 1: culturable heterotrophic bacteria Fig. 8 Experiment 2: culturable heterotrophic bacteria with new phytoplankton and zooplankton to approximately the original densities. 4.3.2. Results and discussion. Similar to previous tests, control groups with and without bisulfite had no decrease in bacteria levels or large changes in phytoplankton or zooplankton. Similar to previous tests, the TRO level dropped approximately 0.5 ppm in the first 5 hours after treatment, dropped approximately 1.0 ppm during the first 24 hours, and then slowly decreased. In AUGUST 2006 mesocosms with added bisulfite, TRO dropped immediately from 2.5 ppm to 0 ppm. Bacteria counts were greatly reduced in the treatment tanks by 5 hours compared to control tanks. The mortality pattern was identical, as shown in Fig. 2, as long as there was a residual TRO present in the tank. Those tanks that were neutralized in Group B (TRO at 0.0 ppm) showed immediate bacteria growth to control levels that were sustained for the remainder of the test. This suggests that the dechlorinated water was no longer toxic or inhibitory JOURNAL OF SHIP PRODUCTION 169 Table 3 Trihalomethane results Trihalomethanes (g/L) Time (h) 0 0 24 24 27 48 48 Treatment Chloroform Bromoform Bromodichloromethane Chlorodibromomethane Total THM (TTHM) Control Treatment B Treatment A Treatment B Treatment B Treatment B Treatment A Detection limits <0.5 <0.5 <0.5 NA <0.5 <0.5 <0.5 0.5 <0.5 43.1 67.7 NA 70.6 68.1 68.0 0.5 <0.5 <0.5 <0.5 NA <0.5 <0.5 <0.5 0.5 <0.5 1.1 1.8 NA 1.9 1.8 2.2 0.5 <0.5 44.2 69.5 NA 72.5 69.9 70.2 0.5 NA ⳱ not analyzed. Table 4 Haloacetic acids results Haloacetic Acids (g/L) Time (h) 0 0 24 24 27 48 48 Treatment DBAA MBAA MCAA DCAA TCAA HAA5 BCAA Control Treatment B Treatment A Treatment B Treatment B Treatment B Treatment A Detection limits <1.0 1.9 1.8 7.8 7.1 6.2* 9.9 1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 1.4 1.0 <2.0 <2.0 <2.0 <2.0 <2.0 <2.0 2.3 2.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 1.6 1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 1.0 <1.0 1.9 1.8 7.8 7.1 6.2 15.2 1.0 <1.0 <1.0 <1.0 <1.0 <1.0 <1.0 9.2† 1.0 DBAA ⳱ dibromoacetic acid; MBAA ⳱ monobromoacetic acid; MCAA ⳱ monochloroacetic acid; DCAA ⳱ dichloroacetic acid (Severn); BCAA ⳱ bromochloroacetic acid; TCAA ⳱ trichloroacetic acid; NA ⳱ no analysis performed. *Outside holding time. † Matrix-induced bias of saltwater. Table 5 Time (hours) 0 0 24 24 27 48 48 Bromate results Treatment (g/L) Control Treatment B Treatment A Treatment B Treatment B Treatment B Treatment A Detection limits <1.0 <1.0 <1.0 <1.0 NA NA <1.0 1.0 to bacteria following the addition of sodium bisulfite. The number of culturable bacteria increased from 10 to 106 CFU/L. After chlorination, chlorophyll a concentrations in treatment tanks ranged from 0.00 to 0.02 g/L, indicating an almost total eradication of phytoplankton resulting from chlorination. Again, the pattern was similar to Fig. 4. For Group B and after neutralization and back addition, the chlorophyll a levels were at and sustained at control group concentrations. This indicated a nontoxic environment for phytoplankton. Zooplankton showed a 98% to 100% mortality rate when treated with 3.5 ppm chlorine compared to controls, similar to Test 2. After dechlorination and restocking of Group B, new zooplankton survived as well as in control samples. 170 AUGUST 2006 Disinfection by-product analysis was performed throughout the mesocosm testing. Although there are no DBP regulations for ballast water, regulated drinking water DBPs and their US EPA maximum contaminant level (MCL) standards were used as a guide for the DBP analysis. Drinking water MCLs for the tested by-products are 0.080 mg/L for the total trihalomethanes (TTHM), 0.060 mg/L for five selected haloacetic acids (HAA5), and 0.010 mg/L for bromate. In Test 3, TTHM and HAA5 levels were below drinking water MCL in all samples. Bromochloroacetic acid is not an HAA5 regulated under current drinking water standards. Results for the various disinfection by-product concentrations are listed in Tables 3, 4, and 5. The results indicate that the DBP created in treating ballast water are below the limits for all three components of disinfection by-products for drinking water. 5. Conclusions Several conclusions can be made from the research completed by the University of Washington: 1. The BALPURE Electrooxidation System, when used to treat incoming seawater, is a promising treatment technology to meet the proposed IMO standards. 2. Dechlorinated ballast water effluent (discharge) was not toxic to Puget Sound organisms. 3. Disinfection by-products that were formed during the tests JOURNAL OF SHIP PRODUCTION were below the concentrations set for drinking water standards. 4. Filtration was not necessary to meet IMO standards if sufficient TRO was added and remained. 5. Operating costs are less than $0.02 per m3 of ballast water based on power and sulfite requirements. References CANADIAN COAST GUARD. 1992 A review and evaluation of ballast water management and treatment options to reduce the potential for the introduction of nonnative species to the Great Lakes, Pollutech Environmental Limited for the Canadian Coast Guard, Ship Safety Branch. DICKMAN, M., AND ZHANG, F. Z. 1999. Mid-ocean exchange of container vessel ballast water. 2: Effects of vessel type in the transport of diatoms and dinoflagellates from Manzanillo, Mexico, to Hong Kong, China, Marine Ecology Progress Series, 176, 253–262. AUGUST 2006 View publication stats HOLM-HANSEN, O., AND RIEMANN, B. 1978 Chlorophyll a determination: improvements in methodology, Oikos, 30, 438–448. IMO. 2004 Ballast Water Management Convention, International Maritime Organization. MONTANI, S., MEKSUMPUN, S., AND ICHIMI, K. 1995 Chemical and physical treatments for destruction of phytoflagellate cysts, Journal of Marine Biotechnology, 2, 179–181. NATIONAL RESEARCH COUNCIL. 1996 Stemming the Tide: Controlling Introductions of Nonindigenous Species, Committee on Ships’ Ballast Operations, Marine Board, Commission on Engineering and Technical Systems, National Academy Press, Washington, DC. PARSONS, T. R., MAITA, Y., AND LALLI, C. M. 1984 A Manual of Chemical and Biological Methods for Seawater Analysis, Pergamon Press, New York. RIGBY, G. R., HALLEGRAEFF, G. M., AND SUTTON, C. 1999 Novel ballast water heating technique offers cost-effective treatment to reduce the risk of global transport of harmful marine organisms, Marine Ecology Progress Series, 191, 289–293. XIE, Y. 2000 Disinfection by-product analysis in drinking water, American Laboratory Shelton, 32, 50–54. JOURNAL OF SHIP PRODUCTION 171