**2. Objectives**

Environmental impacts from the operations of intakes have been studied throughout the literature. The literature, however, does not include a comparison of the impact assessment methods. The objectives of this paper are:

	- o assist with design of intakes
	- o delineate zone of influences
	- o quantify impacts

While impingement risk was assessed for the case study discussed here, this paper focuses on the entrainment element. Entrainment was considered in two different contexts: (1) entrainment of non-mobile IP based on concentrations in the water column and (2) entrainment of mobile life stages based on a zone of influence and burst swimming speeds. Since the ballast water along with any entrained organisms is removed from the area by the vessel, entrainment mortality was assumed to be 100%.

To accomplish these objectives, two 3-dimensional models were used. The first 3-dimensional model was used to develop the flow fields and was based on Delft3D [1,2]. The second model was based on Generalized Environmental Modeling System for surface waters [3–5] and was used to estimate the entrainment of IP from an estuarine environment during the short-term operation of a ballast water intake for an LNG carrier. The equivalent adult model (GEMSS-EAM) assesses the equivalent adults lost through entrainment of ichthyoplankton using the modeled flow fields and incorporates four different methods developed between 1978 and 2005. These methods stem from the same approach of considering lifestages, lifestage durations, natural mortalities related to planktonic lifestages and entrainment mortalities attributable to the intake operations. The study also considers the potential for juveniles and adults to become entrained based on their swimming ability relative to the zone of influence at the intake. The study also considers the uncertainty in estimates for the lifestage data and, as such, performs sensitivity analyses to evaluate the confidence level achievable in such quantitative estimates for entrainment.

For mobile life stages, the focus was on the endangered shortnose sturgeon (*Acipenser brevirostrum*) and the protected (due to its similar appearance) Atlantic sturgeon (*Acipenser oxyrhynchus*). The second element of the analysis considered the burst swimming speeds of adults and juveniles and their ability to escape either impingement (larger adults) or entrainment (smaller adults and juveniles).

All entrainment calculations relied on a hydrodynamic model for the estuary that simulates a 3-dimensional near-field zone of influence in a tidal environment. The 3-dimensional hydrodynamic model was run to predict baseline (no intake operation) hydrodynamic conditions which were contrasted against the conditions during the intake operations. The two modeled conditions were quantitatively compared and the resulting changes in the hydrodynamic conditions were calculated. For plankton, the model was used to predict the number of each species group and planktonic life stage entrained as a function of the volume of water and concentration of IP by species group and life stage. This analysis was then introduced to an Equivalent Adult Model (EAM) to project the number of adults lost to the population. For mobile life stages, the changes were quantitatively compared against swimming speeds for shortnose sturgeon (*Acipenser brevirostrum*) and Atlantic sturgeon (A*cipenser oxyrhynchus*). Sustained, prolonged and burst speeds were considered during evaluation of effects of the intake zone of influence on these species and their ability to escape entrainment.

#### **3. Equivalent Adult Model (EAM)**

The GEMSS-EAM model was developed by extending the Equivalent Adult Model (EAM) formulation described in Horst [6] and Goodyear [7] to include the entrainment estimates from near field regions of a surface water intake. The Goodyear EAM [7] estimates the numbers of adult fish that would result based on the early life-stage population. EAM estimates the loss of IP vulnerable to intake system withdrawals based on the fraction of water volume drawn from various areas throughout the range of vulnerability multiplied by the density of vulnerable IP in those areas. The analysis allows direct comparison of what mortality occurs as a result of fish losses due to entrainment and impingement, compared to populations in the area. From this point, the IP losses are projected to equivalent adults using four methods available in the literature:

Method 1. Extension of the Horst [6] and Goodyear [7] methods to near-field entrainment estimates; Method 2. Adjustment of the Horst [6] and Goodyear [7] based estimates to consider natural mortality;

Method 3. Adjustment using a survival function developed by EPRI [8]; and

Method 4. Adjustment using a survival function published by Exponent [9].

These methods provide a range of projected equivalent losses which can be used to assess the impacts of intake operations.

The hydrodynamic and entrainment modeling is used as input densities of IP from the field sampling data from the sampling regions. The density of a species for a given life-stage "j" at the intake during week "i" is defined mathematically as DWi,j. If Di,j,k is the density of the same species for life stage "j" and in region "k" during week "i", then its density at the intake becomes:

$$DW\_{i,j} = \sum\_{k} D\_{i,j,k} \to\_{i,k} \int E\_{ol,k} \tag{1}$$

Where Ei,k/EOi,k is the fraction of density entrained from a sampling region to the initial sampling region density. The sampling density is population density which is the number of organisms per 1000 ft³ of water and is defined individually for each species, lifestage, region and week. Thus, using the parameters computed from the hydrodynamic and transport model, the intake density required for performing the entrainment analysis leading to the EAM evaluation, DWi,j can be estimated from the observed densities, Di,j,k in all sampling regions, weighted by the fractional entrainment from that region.

#### *3.1. EAM Modeling from Biological Sampling Region Data*

Placing the fractional loss of IP data into context is difficult, largely due to incomplete knowledge of the total population of vulnerable life stages available, their period of vulnerability within each age class or life stage (to the point where they are mobile and can avoid the effects of entrainment or impingement), or their mortality through the system (assumed to be 100%: all entrained IP are lost). The model calculates the portion lost due to entrainment based on the hydrodynamic modeling and the population of vulnerable IP throughout the area potentially affected by the water withdrawals. The identification of vulnerable IP population was based on the extent of the withdrawal modeled by the hydrodynamic model.

Having calculated the number of IP lost, the next task is to place that loss into some context, such as the equivalent number of adults lost to the system, as a result of losses at each life stage (day-age-class, cohort, *etc.*). There are two methods to complete that analysis, either estimating the population of adults that loss would have represented by forecasting the numbers of adults that would have resulted from the population of earlier lifestages by cascading throughout successive lifestages to the adult stage, or estimating the stock size from which those numbers derived in the first place by hind-casting losses of IP into adults that would have produced that number of those life stages or age-classes of fish. Forecasting the loss of adults applies some estimate of survival to adulthood for each life stage, which itself is subject to debate. We address the former, using the EAM tool.

For a species, the number of equivalent adults entrained can be expressed mathematically as:

$$Na = \sum\_{j=1}^{n} N\_{~j} S\_{~j}^{\*}\tag{2}$$

where *Nj* is the number of individuals at life stage (or age class) "j" that have been entrained, *Sn <sup>j</sup>* is the survival probability from the jth life stage class to adulthood, and "n" is the total number of life stages before reaching adulthood. Equation (2) can be expanded to give the number of organisms entrained from each sampling region for each weekly set of observed densities, or *Ni,j,k*. It can be computed as:

$$N\_{i,j,k} = D\_{i,j,k} S\_j^\* V\_{i,k} F E\_{i,k} \tag{3}$$

where *FEi,k* is the fraction entrained from region "k", and is computed from the hydrodynamic and transport model as:

$$FE\_{\iota,k} = E\_{\iota,k} \mid E\_{\iota\iota,k} \tag{4}$$

The net equivalent adults entrained, *Na*, from all the regions and over the entire simulation period, then, is:

$$Na = \sum\_{i} \sum\_{j} \sum\_{k} N\_{i,j,k} \tag{5}$$

Expanding Equation (5) using Equations (3) and (4) gives:

$$\mathbf{Method 1} \colon Na = \sum\_{i} \sum\_{k} \left\{ \sum\_{j} D\_{i,j,k} S\_j^{\pi} \right\} FE\_{i,k} V\_{i,k} \tag{6}$$

This formulation of the EAM, with application over several sampling regions, using weekly density data and results from the hydrodynamic model provides a necessary and important refinement to its computation as recommended by Horst [6].
