The study reported herein evaluated three ballast-treatment technologies—ultraviolet radiation (UV), thermal (heat), and acoustic (ultrasonics)—for potential application to secondary treatment of ballast water (refer to Table 1-1). The study was sponsored by the Northeast-Midwest Institute (NEMW) and the Lake Carriers’ Association (LCA) under the Great Lakes Ballast Technology Demonstration Project (http://www.nemw.org/ERballast.htm).

Vendors and manufacturers (hereafter collectively referred to as "vendors") of water-treatment technologies contacted during this study were asked to provide data on their current product lines and on products they are developing. Contacted vendors were identified through:

• Professional contacts of Northeast-Midwest Institute and Battelle staff;
• Library and worldwide-web internet searches;
• Researchers, vendors, and regulatory agencies at nonindigenous species conferences and meetings; and
• Trade publications and associations of the wastewater, drinking water, and aquaculture industries.

Although most of the useful information obtained during the study came from direct telephone surveys with treatment vendors in the wastewater and drinking water industries, valuable information and leads were obtained from each of the above-listed sources.

Many of the 15 vendors who provided information to this study, several sold only made-to-order equipment or did not market off-the-shelf equipment that could meet the parameters for ship use. In these cases, Battelle researchers asked for calculated technical data from the vendors. Vendors were asked to predict sizes, operational requirements, and performance specifications of their treatment systems if they were modified for shipboard use or scaled up to the necessary flow rates. In cases were vendor-calculated data were unavailable, Battelle researchers made estimations, if possible. All cases where Battelle researchers calculated technical data involved vendor equipment that operated at lower than required flow rates (i.e., <1,200 or 8,000 gpm). In these instances, it was assumed that low-volume equipment units could be plumbed in parallel configurations to achieve the 1,200 to 8,000 gpm rates, and that equipment dimensions and power requirements would be linear.² Battelle researchers note, however, that scaled-up data for low-volume equipment is probably conservative. There will likely be significant economies realized if low-volume equipment is scaled-up in the design or manufacturing processes, or if mixes of different sized units are installed on a vessel to optimize treatment capacity.

Battelle did not manipulate biological performance data provided by the vendors. Much of these data, usually extracted from marketing materials, was neither verifiable nor comparable across vendors, particularly among the three technologies. Thus, biological performance data was accepted as accurate, and used in the data analysis at the discretion of the researchers.

Wherever possible, equivalent technical and performance data were compared in this study of secondary treatment technologies. Technical data that were scaled by either the vendors or Battelle were assumed to add uncertainty relative to quantifying treatment-equipment dimensions, weights, and power requirements. However, any generated inaccuracies are probably conservative with respect to operational requirements, and are unlikely to affect comparison of biological performance of the treatment equipment. Problems with accuracy and comparability among the biological performance data were handled by best professional judgement of the researchers.

Readers of this report should recognize that both ballast management and water treatment are rapidly advancing fields. The information presented in this report is a "snapshot" of the three studied technologies. Most of the vendors surveyed during the study are producing products for the highly competitive industrial-waste and municipal water treatment markets. The information that they provided to Battelle intentionally emphasized the positive attributes of their products. Despite this self-promotion, it was evident that all of the contacted vendors are continually improving their products. The equipment that they are currently marketing will likely become smaller, lighter, more efficient, and less expensive as the technological advances are made. Most vendors expressed certainty that once exact shipboard specifications are determined, an efficient treatment system with their technology could be designed for onboard ballast-water treatment.

The focus of this study was on the biological effectiveness of secondary-treatment systems for ballast water. Effectiveness was defined by a technology’s ability to disinfect ballast water of microorganism communities. Full disinfection of ballast water was defined as inactivation, destruction, and/or removal of bacteria, viruses, cysts, and other aquatic microorganisms for making the water "safe" for discharge in cargo-port waters. In this study, this level of disinfection is destruction of 99.9% of the viable microorganism population (a 3log reduction) or better.

Microbiologists concerned with the rate of killing of bacteria sometimes refer to "Chick’s law for killing," which hypothesizes that the rate of killing of bacteria follows first order kinetics when plotted exponentially (microbiologists only think in exponents as the numbers of microorganisms per sample are so high, in the order to e+4 to e+8 per milliliter in the real world). Roughly translated, any ionizing radiation, or heat, or certain chemicals will kill an individual bacterial cell with a single hit. However, due to the number, concentration, and size of small microorganisms (e.g., bacteria); and the intensity and size of photons, electrons, or chemical molecules; a straight line is generated when the log number of bacteria is plotted against the time of exposure. However, microbiologists have found this "law" to be only partially correct.

In treatment systems short of sterilization,³ surviving microorganisms will be more resistant to the successive treatments by the same agent. Resistance to a killing agent (chemicals, UV, heat, and ultrasonics) is most pronounced if exposure is low and over a prolonged period. Resistance development is generally not a concern in single-treatment ballast-water systems, especially if the killing agent is sufficiently strong and exposure is short duration. Resistance development could be of concern, however, if successive treatments of ballast water are considered (e.g., treating water on both intake and discharge).

Particles, films, and crevices are significant obstacles to killing microorganisms. Academic microbiologists generally only work with cultures in dissolved media, so that microorganisms are the biggest particles in the solutions. In real-world systems, anything bigger then a microorganism can potentially block the killing agent from getting to the microorganism. One classic example is a microbial film or "slime layer," where treatment of the outside of the layer with UV or heat has virtually no affect on microorganisms on the protected side of the layer. Barriers to chemical or physical killing agents are thus of major concern to the design and operation of a ballast-water treatment system.

In summary, it is important to understand that the 3log disinfection level used as a measurement standard for biological effectiveness is somewhat arbitrary. With the very large volumes of water being quickly treated in shipboard systems, 3log disinfection means that many individual organisms (0.01% of the original population) will survive and potentially invade a new environment. However, readers should also realize that 3log disinfection is a substantial change. In other treatment processes (e.g., food and beverages), a 3log disinfection is considered "sterile." In a practical sense, the remaining numbers of microorganisms are generally too low to cause spoilage within normal holding times for milk, water, or other material. Furthermore, natural defense systems (e.g., in the human body) can overcome small numbers of disease organisms. Analogously, receiving waters are the human body. Released microorganism populations that have undergone 3log disinfection are considered by the authors of this report to have low potential for successfully invading a new environment.

Organism and Sediment Size vs. Treatment. The relationship of target organism size and suspended sediment size is very important relative to the primary and secondary ballast-water treatment. This study evaluated secondary treatment efficacy under no-, 250-m m-, 100-m m-, and 25-m m primary treatment filtration scenarios.⁴