Water column. Characterization factors for suspended particles (p) in the water column (CFp,sea,turb,sea) are predicted analogous to CFs for chemicals, as a function of the fate factor (FFp,e,j) and the effect factor (EFp,i,j) (Eqn. 1). The fate factor for suspended particles represents the marginal increase in mass of suspended particles in the marine water column (j = sea) due to a marginal increase in particle emissions to seawater (e = sea). This fate factor can be approximated by the residence time of the particles in the marine water column. The residence time of particles in the marine water column is related to the depth of the water column (i.e., point of discharge) and the net falling velocity of suspended particles (Eqn. 2), the latter of which is a function of the particle diameter (Rye et al. 2008; Supplemental Information).
where hw = (average) depth of the water column (m); d = particle diameter (m); g′ = reduced gravity: g·((ρparticle – ρseawater)/ρseawater) (m/d2); g = standard gravity (9.813 m/s2); ρparticle = density of particle (2.5 × 103 g/L) (Rye et al. 2006); ρseawater = density of seawater (1.0 × 103 g/L); ν = kinematic viscosity (1.358 × 10−6 m2/s at 10 °C for seawater) (m2/d) (Rye et al. 2006); and K = empirical constant (1.054) (sans unit) (Rye et al. 2006)
The effect factor for turbidity due to suspended particles in the seawater column (EFp,turb,sea) can be calculated similar to effect factors for toxic stressors (Eqn. 3).
where EFp,turb,sea = the effect factor for turbidity in the water column for particles (p) (PAF × m3/kg); and HC50, p,turb = hazardous concentration at which 50% of the species are exposed above their EC50p,turb (g/L).
We obtained acute effect data for barite particles and bentonite particles from Smit et al. (2008). For barite particles, acute effect data is available for 15 species, including zooplankton, algae, mollusks, crustacea, oligochaetes, and fish. For bentonite particles, acute effect data is available for 12 species, including zooplankton, algae, mollusks, crustacea, and fish. We combined the effect data of barite and bentonite to obtain and average (acute) HC50, p,turb, which can be applied to other suspended particles of a similar diameter range, such as small cutting particles. The average (acute) HC50, p,turb is 2.3 g/L. An acute to chronic ratio of 2 is applied to extrapolate data to chronic effect data (Rosenbaum et al. 2008). This results in a (chronic) effect factor of 4.3 × 10−1 PAF·m3/kg.
The overall impact of particle deposition on bottom-dwelling biological communities is a function of 3 dynamic, interacting processes (Bolam and Rees 2003; Breuer et al. 2004; Smit et al. 2008):
Deposit layer (cutting pile) characteristics such as formation rate, frequency of discharge, surface area, thickness, porosity, and grain size
Initial benthic response
Benthic recovery rate
Cutting pile characteristics vary from site to site, depending on local environmental conditions (water depth, current velocity, turbulence) and discharge characteristics such as cuttings properties (e.g., size distribution, density, and sphericity), drilling mud applied (e.g., size distribution and agglomeration potential), and platform configuration (discharge rate and discharge depth; Neff et al. 2000; Breuer et al. 2004). The Norwegian Sea and the central and northern North Sea are characterized by relatively weak currents. Here, discharged cuttings tend to flocculate and accumulate under and around platforms, resulting in the formation of so-called cutting piles (Breuer et al. 2008). Once deposited, these cuttings become consolidated and more resistant to erosion processes, and in situ transport of cuttings is rare (Breuer et al. 2008). In areas with shallow water depths and higher tidal flows, such as the southern North Sea, no extant cutting piles can be identified (Neff et al. 2000; Breuer et al. 2008).
Effects on benthic organisms are considered to fall into 2 relatively distinct categories: initial (short-term) effects due to burial by drilling fluid and cuttings and longer-term effects due to physical (textural) alteration of the sediments (USEPA 2000). The initial effects of burial of species appear to be determined by the depth of burial; burial rate; tolerance of species, such as life habitats, escape potential, and low O2 tolerance; nature of deposited material, such as grain size different from the native sediment; and season, e.g., mortality rate by burial is higher in summer than winter (Smit et al. 2008). The deposition of drill cuttings changes the texture and structure of the sediment, rendering the substrate less suitable as habitat for some benthic species and more suitable for others (Smit et al. 2008). Bioturbation, physical mixing (e.g., by currents), and sedimentation of fresh material results in restoration of original sediment characteristics and eventually the deposited cutting layer is recolonized by benthic organisms. In general, benthic recovery after a disturbance occurs in 3 stages of successional recolonization: 1) the benthos is barren of invertebrates, 2) the benthos is recolonized by opportunistic taxa and faunal abundance increases exponentially, and 3) opportunistic species are replaced by larger, slower-growing “equilibrium species” (Blanchard and Feder 2003). Recovery is considered to be complete when a community varies similarly to predisturbance conditions or to the ambient community in a nearby location. In addition, a completely recovered community is able to maintain itself and returns to greater than 80% of the original diversity and biomass (Essink 1999; Blanchard and Feder 2003). Large differences in recovery rates are observed (Bolam and Rees 2003; Neff 2000). These differences have been attributed to the number of successional stages required to regain the original community composition and environmental characteristics (Bolam and Rees 2003). If the perturbation destroys the original community and recovery starts from totally defaunated sediments, the recovery rate is related to the ability of the surrounding, undisturbed community to supply migrating adults and/or recruiting larvae (Bolam and Rees 2003). In general, relatively unstressed marine environments (deep, physically stable habitats experiencing infrequent sediment movement from wave action) take between 1 to 4 y to recover, whereas more naturally stressed areas (shallow estuarine habitats, which are physically transient), recover within 9 months, because species are more adapted to unpredictable conditions (Bolam and Rees 2003).
The CF for burial quantifies the impact of particle deposition on the seafloor, as a function of the initial cutting pile height, the initial benthic response, and the benthic recovery rate (Eqn. 4).
where CFp,sea,bur,sed = characterization factor for burial of benthic communities due to particle deposition on the seabed after a particle emission to seawater (PAF·m3/d/kg); Hb,sed = vertical increase in sediment top-layer (m) at location x,y at time t = 0; f(PAF)Hb,sed = PAF-based response function for burial of species (PAF m−1); Mp,sea = particle emission to seawater (1.0 × 106 kg); RT = recovery time (days) (1460 days); Ip,sea,burial,sed = volume-integrated, total PAF of species at time t = 0 due to particle deposition on the seabed (PAF m3).
The Dose-related Risk and Effect Assessment Model (DREAM) was used to predict the initial burial height (Hb,sed) and impacted surface area (x,y) of the cutting pile, as a function of ocean currents, turbulence, and particle size (Rye et al. 2008). Cutting layer characteristics were predicted for the drilling of a generic well located in the northern part of the North Sea with a total cutting release of 1.0 × 103 tonnes (see Supplemental Data). The initially potentially affected fraction of species was estimated with a PAF-based response function for burial of benthic fauna by particles, established by Smit et al. (2008). This PAF-based response function is a log-normal distribution expressing the PAF of species per centimeter increase in burial layer and is based on chronic effect data for 32 species, including 24 mollusks, 5 crustacea, and 3 polychaetes (Smit et al. 2008). The characterization factor describes the volume of sediment that is impacted by an increase in sediment top-layer due to deposition of cuttings on the seafloor. Benthic fauna is assumed to be concentrated in the first 5 cm of sediment, i.e., z = −0.05 m (Dauwe et al. 1998; Flach et al. 2002). The volume of impacted sediment thus equals the deposition layer surface area (x,y) multiplied with the vertical dimension of benthic species occurrence (z). The recovery rate of 4 years (1460 days) is based on a review study by Neff et al. (2000) on fate and effects of SBM cuttings. Neff et al. (2000) suggested that recolonization by opportunistic species occurs well within 1 y, whereas full recovery of the impacted site takes between 3 to 5 y. We assume a linear recovery rate of the benthic community until full recovery is obtained.