Optimal siting of rainwater harvesting systems for reducing combined sewer overflows at city scale.

The installation of green infrastructure (GI) is an effective approach to manage urban stormwater and combined sewer overflow (CSO) by restoring pre-development conditions in urban areas. Research on simulation-optimization techniques to aid with GI planning decision-making is expanding. However, due to high computational expense, the simulation-optimization methods are often based on design storm events, and it is unclear how much different rainfall scenarios (i.e., design storm events vs. long-term historical rainfall data) impact the optimal siting of GI. The Parallel Pareto Archived Dynamically Dimensioned Search (ParaPADDS) algorithm in a novel simulation-optimization tool OSTRICH-SWMM was used to leverage distributed computing resources. A case study was conducted to optimally site rainwater harvesting cisterns within 897 potential subcatchments throughout the City of Buffalo, New York. Seven design storm events with different return periods and rainfall durations and a one-month historical rainfall time series were considered. The results showed that the optimal solutions of siting cisterns using event-based scenarios, though less computationally expensive, may not perform well under continuous rainfall scenarios, suggesting design rainfall scenarios should be carefully considered for optimizing GI planning. The impact of rainfall scenarios was particularly significant in the middle region of the Pareto front of multi-objective optimization. Utilizing high-performance parallel computing, OSTRICH-SWMM is a promising tool to optimize GI at large spatial and temporal scales.

Hydrologic evaluation of a poplar phytoextraction system.

This paper reports the results of a seven-year study of a phytoremediation system consisting of 349 hybrid poplars (Populous deltoids x Populus nigra 'DN-21') that were emplaced to reduce the migration of impacted groundwater toward a nearby stream. For the relatively small tree stand, the achievement of hydraulic containment was challenged by several factors, including topographic constraints on tree placement, spatial heterogeneity in depth-to-groundwater, and lack of controls on surface runoff. However, the interpretation of measured diurnal water table fluctuations indicated substantial and increasing levels of groundwater extraction, consistent with poplar sap flow measurements collected during two growing seasons. The project also provided unique educational experiences for undergraduate and graduate students from multiple academic disciplines who participated in monitoring, site maintenance, and data interpretation. Novelty statement The paper adds to the limited literature on poplar-based phytoextraction of contaminated groundwater by describing a seven-year case study with detailed hydrologic measurements. Novel elements include (1) a very large data set of poplar sap flow measurements that provides insight into potential spatial and temporal variability, and (2) a detailed application of an under-utilized analytical tool (water table fluctuations) to evaluate hydraulic capture.

Parameterizing sorption isotherms using a hybrid global-local fitting procedure.

Predictive modeling of the transport and remediation of groundwater contaminants requires an accurate description of the sorption process, which is usually provided by fitting an isotherm model to site-specific laboratory data. Commonly used calibration procedures, listed in order of increasing sophistication, include: trial-and-error, linearization, non-linear regression, global search, and hybrid global-local search. Given the considerable variability in fitting procedures applied in published isotherm studies, we investigated the importance of algorithm selection through a series of numerical experiments involving 13 previously published sorption datasets. These datasets, considered representative of state-of-the-art for isotherm experiments, had been previously analyzed using trial-and-error, linearization, or non-linear regression methods. The isotherm expressions were re-fit using a 3-stage hybrid global-local search procedure (i.e. global search using particle swarm optimization followed by Powell's derivative free local search method and Gauss-Marquardt-Levenberg non-linear regression). The re-fitted expressions were then compared to previously published fits in terms of the optimized weighted sum of squared residuals (WSSR) fitness function, the final estimated parameters, and the influence on contaminant transport predictions - where easily computed concentration-dependent contaminant retardation factors served as a surrogate measure of likely transport behavior. Results suggest that many of the previously published calibrated isotherm parameter sets were local minima. In some cases, the updated hybrid global-local search yielded order-of-magnitude reductions in the fitness function. In particular, of the candidate isotherms, the Polanyi-type models were most likely to benefit from the use of the hybrid fitting procedure. In some cases, improvements in fitness function were associated with slight (<10%) changes in parameter values, but in other cases significant (>50%) changes in parameter values were noted. Despite these differences, the influence of isotherm misspecification on contaminant transport predictions was quite variable and difficult to predict from inspection of the isotherms.

Isotherm ranking and selection using thirteen literature datasets involving hydrophobic organic compounds.

Numerous isotherm expressions have been developed for describing sorption of hydrophobic organic compounds (HOCs), including "dual-mode" approaches that combine nonlinear behavior with a linear partitioning component. Choosing among these alternative expressions for describing a given dataset is an important task that can significantly influence subsequent transport modeling and/or mechanistic interpretation. In this study, a series of numerical experiments were undertaken to identify "best-in-class" isotherms by refitting 10 alternative models to a suite of 13 previously published literature datasets. The corrected Akaike Information Criterion (AICc) was used for ranking these alternative fits and distinguishing between plausible and implausible isotherms for each dataset. The occurrence of multiple plausible isotherms was inversely correlated with dataset "richness", such that datasets with fewer observations and/or a narrow range of aqueous concentrations resulted in a greater number of plausible isotherms. Overall, only the Polanyi-partition dual-mode isotherm was classified as "plausible" across all 13 of the considered datasets, indicating substantial statistical support consistent with current advances in sorption theory. However, these findings are predicated on the use of the AICc measure as an unbiased ranking metric and the adoption of a subjective, but defensible, threshold for separating plausible and implausible isotherms.

Statistical techniques for analyzing of soil vapor intrusion data: a case study of manufactured gas plant sites.

UNLABELLED: As part of an ongoing study of soil vapor intrusion (SVI), concentration data for approximately 2000 air and vapor samples were assembled from remedial site investigations and stand-alone assessments conducted at New York State Manufactured Gas Plant (MGP) sites. Vapor samples were collected from ambient outdoor air indoor air, beneath building slabs, and from outside of buildings. Despite the large sample size, the considerable variability in compound and sample-specific censoring limits inhibited the use of conventional tools for statistical interpretation. This paper describes the development and application of improved statistical tools to address an unusually high degree of data censoring and possible artifacts related to uneven distributions of samples across sites and buildings. In addition to methods for calculating population percentiles and associated confidence intervals, methods for comparing the population of MGP-SVI data with a reference population were also developed and evaluated via illustrative comparisons with the published 2001 EPA Building Assessment Survey and Evaluation (BASE) study of industrial buildings. The focus of this work is on the development and evaluation of new statistical methods; a more complete summary and evaluation of the full NYS MGP-SVI data set will be presented in a companion paper. IMPLICATIONS: Data from vapor intrusion and other environmental studies are often stratified and/or censored, which complicates comparisons with background data or reference populations. In some cases, statistical methods for censored data can be modified to support population-based inference and reduce biases associated with the presence of repeated measurements from multiple sources. Such modifications are particularly appropriate for retrospective data mining studies that are not guided by a formal experimental design.

Natural zeolite permeable treatment wall for removing Sr-90 from groundwater.

Experimental and modeling studies were completed to investigate the potential performance of a sorbing permeable treatment wall (PTW) comprised of natural zeolite for removal of strontium-90 (Sr-90) from groundwater at the West Valley Demonstration Project (WVDP) near Buffalo, NY. Multiple column tests were performed at the University at Buffalo (UB) and WVDP for periods ranging from 6 months to 2 years; UB columns were supplied with synthetic groundwater referenced to anticipated field conditions, while radioactive groundwater obtained on site was used for the WVDP columns. The primary focus was on quantifying the competitive cation reactions among five cations (Na(+), K(+), Ca(2+), Mg(2+), Sr(2+)) and Sr-90 with data obtained from the column studies used to estimate Gaines-Thomas (GT) selectivity coefficients. The resulting six-solute transport model provided flexibility to explore the influence of PTW parameters on long-term PTW performance, including variations in Sr-90 concentrations and groundwater geochemistry. The natural zeolite PTW is a viable method for in situ removal of Sr-90 from groundwater and potentially applicable to other sites contaminated by Sr-90.

Application of heuristic optimization techniques and algorithm tuning to multilayered sorptive barrier design.

Although heuristic optimization techniques are increasingly applied in environmental engineering applications, algorithm selection and configuration are often approached in an ad hoc fashion. In this study, the design of a multilayer sorptive barrier system served as a benchmark problem for evaluating several algorithm-tuning procedures, as applied to three global optimization techniques (genetic algorithms, simulated annealing, and particle swarm optimization). Each design problem was configured as a combinatorial optimization in which sorptive materials were selected for inclusion in a landfill liner to minimize the transport of three common organic contaminants. Relative to multilayer sorptive barrier design, study results indicate (i) the binary-coded genetic algorithm is highly efficient and requires minimal tuning, (ii) constraint violations must be carefully integrated to avoid poor algorithm convergence, and (iii) search algorithm performance is strongly influenced by the physical-chemical properties of the organic contaminants of concern. More generally, the results suggest that formal algorithm tuning, which has not been widely applied to environmental engineering optimization, can significantly improve algorithm performance and provide insight into the physical processes that control environmental systems.

Performance assessment of a zeolite treatment wall for removing Sr-90 from groundwater.

Laboratory and modeling studies were conducted to assess the potential performance of a permeable reactive barrier constructed of a natural zeolite material at the West Valley Demonstration Project in western New York State. The results of laboratory column tests indicated that the barrier material would be effective at removing strontium from groundwater under natural gradient conditions. Two one-dimensional contaminant transport models were developed to interpret the data. A single-solute retardation factor model provided good agreement with the column test data, but time-consuming extraction and analysis of the zeolite material was required to parameterize the model. A preliminary six-solute model was also developed based on the assumption of competitive cation exchange as the primary removal mechanism. Both models yielded similar predictions of the long-term performance of the barrier, but the cation exchange model predicted higher effluent concentrations during the first 1000 pore volumes of operation. The cation exchange framework has several advantages, including the ability to calibrate the model using only data from column effluent samples, and the ability to account for site-specific differences in the groundwater cation composition. However, additional laboratory work is needed to develop a suitably robust model.