Fats, oils, and grease (FOG) wastewater is a kind of high-strength industrial wastewater that contains a high amount of total suspended solids (TSS), chemical oxygen demand (COD), and proteins. One industry that has high-FOG wastewater is the rendering industry, which converts animal byproducts to protein meal and fat, which have value as a commodity. FOG wastewater in the rendering industry is usually treated by dissolved air flotation (DAF) and the recovered proteins and fats are sent back to the rendering process stream. An important drawback to DAF is that chemical flocculants are usually needed, but these chemicals become contaminants in the final protein and fat streams sold to customers. Another drawback is that oxygen in the DAF process can oxidize proteins and degrade their quality.

An alternative to DAF is membrane technology. Because membranes have not been heavily used in FOG wastewater applications (especially in the rendering industry) this project sought to create a test system to evaluate membranes in this context, with real-world operating conditions.

There were three main objectives in this project. The first objective was to build a field-deployable semi-autonomous filtration unit. This was achieved by improving an existing system from previous research. Other pumps, valves, and electronic components were added to the system to realize semi-autonomous filtration. A control program was created in LabVIEW for system operation.

The second objective was to operate the system continuously to test its capabilities. The algorithms in the control program were updated according to the results from test filtrations of tap water, lake water, and effluent from a wastewater treatment plant.

The third objective was to employ chemical cleaning to recover flux decline, similarly as would be done in a full-scale system. Rendering plant wastewater was used for testing the system and three chemicals were employed for cleaning the foulant formed during filtration.

A software interface was built for system operation and data recording. Three main programs aimed to control the filtration loop, actuator valve, and record data, respectively. Other programs were also created for the stabilization of the system and protecting the hardware. For example, an averaging program was used for decreasing the influence of extreme values. A pump reverse program aimed to protect the flowmeter after the backwash. The hardware cooperated with the software interface for signal processing and fluid handling. A data analysis program was coded in MATLAB for plotting and calculating experimental results.

The ability and stability of the system were tested. Results showed that the system could handle filtration, backwash, and chemical cleaning in a time-based operational scheme. Twenty-two data were recorded and emails with data and alarms were sent to students and professors during the experiments. Chemical cleaning efficiencies were calculated based on balance flowrate during filtration. A 2% sodium hydroxide solution had the highest flux recovery (70%) compared to 39% for a 2% solution of hydrogen chloride and 56% for a 0.02% solution of ethylenediaminetetraacetic acid (EDTA).