Filtering by
- All Subjects: water treatment
- Creators: Fox, Peter
- Creators: Chemical Engineering Program
Much of Nepal lacks access to clean drinking water, and many water sources are contaminated with arsenic at concentrations above both World Health Organization and local Nepalese guidelines. While many water treatment technologies exist, it is necessary to identify those that are easily implementable in developing areas. One simple treatment that has gained popularity is biochar—a porous, carbon-based substance produced through pyrolysis of biomass in an oxygen-free environment. Arizona State University’s Engineering Projects in Community Service (EPICS) has partnered with communities in Nepal in an attempt to increase biochar production in the area, as it has several valuable applications including water treatment. Biochar’s arsenic adsorption capability will be investigated in this project with the goal of using the biochar that Nepalese communities produce to remove water contaminants. It has been found in scientific literature that biochar is effective in removing heavy metal contaminants from water with the addition of iron through surface activation. Thus, the specific goal of this research was to compare the arsenic adsorption disparity between raw biochar and iron-impregnated biochar. It was hypothesized that after numerous bed volumes pass through a water treatment column, iron from the source water will accumulate on the surface of raw biochar, mimicking the intentionally iron-impregnated biochar and further increasing contaminant uptake. It is thus an additional goal of this project to compare biochar loaded with iron through an iron-spiked water column and biochar impregnated with iron through surface oxidation. For this investigation, the biochar was crushed and sieved to a size between 90 and 100 micrometers. Two samples were prepared: raw biochar and oxidized biochar. The oxidized biochar was impregnated with iron through surface oxidation with potassium permanganate and iron loading. Then, X-ray fluorescence was used to compare the composition of the oxidized biochar with its raw counterpart, indicating approximately 0.5% iron in the raw and 1% iron in the oxidized biochar. The biochar samples were then added to batches of arsenic-spiked water at iron to arsenic concentration ratios of 20 mg/L:1 mg/L and 50 mg/L:1 mg/L to determine adsorption efficiency. Inductively coupled plasma mass spectrometry (ICP-MS) analysis indicated an 86% removal of arsenic using a 50:1 ratio of iron to arsenic (1.25 g biochar required in 40 mL solution), and 75% removal with a 20:1 ratio (0.5 g biochar required in 40 mL solution). Additional samples were then inserted into a column process apparatus for further adsorption analysis. Again, ICP-MS analysis was performed and the results showed that while both raw and treated biochars were capable of adsorbing arsenic, they were exhausted after less than 70 bed volumes (234 mL), with raw biochar lasting 60 bed volumes (201 mL) and oxidized about 70 bed volumes (234 mL). Further research should be conducted to investigate more affordable and less laboratory-intensive processes to prepare biochar for water treatment.
The model developed simultaneously determines the peak stormwater inflow from watershed parameters and optimizes the size of the basin and the number and depth of dry wells based on infiltration, evapotranspiration (ET), and dry well characteristics and cost inputs. The modified rational method is used for the design storm hydrograph, and the Green-Ampt method is used for infiltration. ET rates are calculated using the Penman Monteith method or the Hargreaves-Samani method. The dry well flow rate is determined using an equation developed for reverse auger-hole flow.
The first phase of development of the model is to expand a nonlinear programming (NLP) for the optimal design of infiltration basins for use with bioretention basins. Next a single dry well is added to the NLP bioretention basin optimization model. Finally the number of dry wells in the basin is modeled as an integer variable creating a MINLP problem. The NLP models and MINLP model are solved using the General Algebraic Modeling System (GAMS). Two example applications demonstrate the efficiency and practicality of the model.