Determination of arsenic concentration and distribution in the Floridan Aquifer System

Abstract

Here we report on the different sampling strategies for almost seven years of sampling rocks/sediments for the determination of As within the Intermediate Aquifer System (IAS) and upper Floridan Aquifer System (FAS), a very large and productive limestone aquifer spanning from Georgia into Florida. In the FAS, As contamination has become a recurring problem during aquifer storage and recovery (ASR), particularly in central and south Florida. To investigate these phenomena, samples from solid drill cores and rock cuttings were collected from the Hawthorn Group, Suwannee Limestone, Ocala Limestone and Avon Park Formation. Samples were taken along drill cores and rock cuttings (referred to as ‘interval’ samples) or from particular drill core sections and rock cuttings (referred to as ‘targeted’ samples) likely to contain elevated concentrations of As as indicated by the presence of pyrite, hydrous ferric oxide, organic matter, clay minerals, fracture surfaces, and high permeable (moldic) zones. Arsenic was present in all of the stratigraphic units at low concentrations, close to the global average for As in limestone of 2.6 mg/kg. The highest As concentration was 69 mg/kg. In all units, however, the average bulk As concentration in the targeted samples was substantially higher than that in the interval samples. Based on direct spot measurements by electron microprobe and indirect calculations, pyrite was identified as the main source of As in the FAS. Concentrations in pyrite ranged from less than 100 mg/kg to more than 11,000 mg/kg. Because pyrite is heterogeneously distributed, both vertically and horizontally in the sampled stratigraphic units, the same was observed for the distribution of As. However, As concentrations generally decreased with depth, i.e., highest As values in the Hawthorn Group and lowest As values in the Ocala Limestone and Avon Park Formation. Compared to pyrite, other trace minerals contained much less As. The average As concentrations of the two types of sample media (solid cores and rock cuttings) were quite similar. These results indicate that if simply the average bulk rock As concentration of a geologic unit is the desired outcome of an investigation, either interval or targeted sampling of rock cuttings, seems to be sufficient. This is particularly important when time and money are a factor. This approach could work equally well for any other trace element. Structural sedimentary information, such as fractures, etc., is likely lost, however, when sampling rock cuttings. Thus, if this information is required, solid core samples need to be collected by hollow core diamond drilling.