119 



Laboratory procedures involving the testing of sediments might also be part of this study. For example, 

 experiments examining growth and fecundity, bioavailability, and the partitioning of contaminants. While EPA 

 protocols should be considered for such work, they were intended to examine the effects of compounds which 

 are not metabolized. This is of concern because many contaminants are metabolized into more toxic materials 

 (metabolites). 



In situ experiments may also need to be designed which use the fouling organisms/communities themselves as 

 the subjects of study of bioaccumulation, bioavailability, and genetic responses. For example, the Shell Oil 

 Company has investigated the bioavailability of mercury from sacrificial anodes into mussels and scallops. While 

 no deleterious effects were observed, very little of this type of work has actually been conducted. 



3.2.3 Production Fields - A Historical Perspective, 

 Concerns, and Recommendations 



Early production fields were designed around central facilities each surrounded by satellites having one or two 

 wells. Potentially, each satellite could have its own "impact" area with accumulations of drilling mud and cuttings, 

 corrosion from anodes, etc. The "central" facility (where "production" takes place) might release significant 

 quantities of produced waters. If studies are to take place in these older areas, it is critical that they be ground- 

 truthed. 



More recent production fields do not typically use the satellite design. There is a single large structure with 

 multiple wells; up to 100 wells may be drilled from a single structure. The products of these multiple well 

 structures are then channeled into a single pipeline feeding into the "trunk" line to shore. The shift towards this 

 design has occurred during the past five years. While it may be easier to examine the latter type of "field," any 

 effort to investigate long-term impacts may necessitate the use of "satellite" fields. 



3.2.4 Recommendations: Analyses, QA, and QC 



The level of analytical precision will be determined by whether a few sites or many sites are chosen and the 

 bound on their respective ranges (i.e., sensitivities required). These analyses may include AAS, ICAP-MS (the 

 method of choice for metals), neutron activation, and X-ray diffraction for mineralogy. It was also agreed that 

 all physical, chemical, biological, and geological samples need to come from the same source. For example, in 

 the CAMP program samples are taken with a Hessler-Sandia box core fitted with 25 subcores so that synoptic 

 measurement can be made for biological, chemical, and sedimentological properties (Montagna, section 2.2.8). 



In any bioassay type testing, efforts must be made to use sensitive life stages (e.g., eggs, embryo, and larvae) and 

 only "realistic" concentrations for exposures. The latter would necessitate any bioassay testing be closely 

 coordinated with fate and transport studies. Such work will also require up front quality assurance/quality 

 control (QA/QC). However, while a central "core" of procedures and standards needs to be established, the 

 overall study design should allow for enough flexibility to modify experiments as needed. 



Finally, all of the information must be pulled together into a Hazard Assessment. Here models will be important 

 because they pull all of the pieces together and while they may not actually represent the real environment, they 

 often do supply something which can be just as important: "what pieces of information are missing?" That is, 

 models may tell you what you don't know. Finally, laboratory testing should be used to verify field observations 

 and data wherever possible. 



3.2.5 Areal Designs for Sampling 



The "matched design" or "paired t-test" whereby several platforms are examined at one point in time was agreed 

 upon as the design most suitable for this type of study. This design involves sampling the parameter(s) of interest 

 along a transect originating from the individual structures and making comparisons between the sampling 

 locations along the transect (i.e., as opposed to examining just a few platforms and following them through time: 

 the current status verses a time series change). 



