AguaClara is a Cornell University organization that is developing sustainable water treatment solutions for small communities. Its research and development team focuses on improving flocculation, and other processes that are used, and making them as cost effective and easy to use as possible. One of the current design focuses is flocculation and the parameters that affect it. Specifically G (1/s), the velocity gradient, G-theta, a dimensionless measure of mixing, and alum dose. This past spring the vertical flow pilot plant team designed and built a hydraulic vertical flow flocculator. The flocculator is a small scale version of a flocculator that has been built in Ojojona, Honduras. It is a square tank that has been divided into three sections and consists of alternating baffles spaced evenly to produce a uniform G of 45 1/s. The flow through the tank is serpentine and is at a flow rate of 114 L/min. The original design of the tank was meant to be easily disassembled, with the baffles and dividers being separate modules from the tank. Sampling and testing of the tank was done through glass tube settlers attached to a pump and fed through turbidimeters. These could be moved to sample at different locations and give a reading of settled water turbidity. When the tank first became operational several problems were discovered with the flocculator and the tube settlers. The biggest of the problems was short circuiting which was discovered by visual observation of the tank as well as head loss measurements. The calculated head loss through the tank was around 11 cm while the original measured head loss was only 2 cm. These short circuits were fixed with Kwik Foam, caulk, and sand in order to seal the gaps and make them water proof. Another problem was divider bowing which when fixed by clamping 2x4s to the edge of the dividers to straighten them out, the head loss rose to 7 cm. It appears with the module design of the baffles that this head loss is the greatest achievable head loss and corresponds to a G of 30 1/s with a G-theta of 22,630. Preliminary testing showed that this design was capable of producing water under 1 NTU with a derived alum dose equation at low turbidities. The third section of the tank produced water a the same or greater turbidity than the other two sections of the tank providing evidence that there could be an optimum G?. Future experiments will look into tapered flocculation and finding an optimal G-theta and G value.