Treated Effluent Reuse for GM Mexico Automotive Assembly Plants

General Motors has manufacturing facilities located in arid regions, where the water is supplied from aquifers with limited capacity. In order to sustain the plant water supply for the lifetime of the facility, water conservation measures and end-of-pipe reuse methods have become critical strategies. Two facilities were selected to address the latter strategy, the GM Mexico Ramos Arizpe Complex and the GM Mexico Silao Assembly Plant. Both of these plants are located in very arid climates with little rainfall. Pilot tests for reusing treated effluent from activated sludge processes were performed at these facilities. Feed to the Ramos Arizpe reuse system was clarified secondary effluent from the industrial biological systems. Feed to the Silao reuse system was clarified effluent from the combined industrial and biological treatment system. The common pilot process consists of continuous microfiltration (CMF) followed by reverse osmosis. A full-scale system is being started up at the Ramos Arizpe Complex and is now in the debugging stage.The average fluxes for the two CMF pilot units were 41 gfd and 35 gfd respectively versus a design flux of 39 gfd for the full scale RAC plant. The run time between chemical cleanings was between 17 and 43 days. The feed and filtrate turbidity for the Ramos Arizpe pilot plant were in the ranges of 5 to 36 and 0.01 to 0.36 NTU, with removal efficiencies greater than 99%. The Silao feed turbidities were similar to the RAC values, but the filtrate turbidity was generally higher, in the range of 0.2 to 0.8 NTU. It is important to note that in Silao, turbidity was measured using a laboratory turbidimeter that usually provides results higher than those of an in-line instrument. Essentially, the COD removal across both pilot CMF units was insignificant, indicating that the majority of the residual COD was not filterable.Trans-membrane pressure and resistance recovery was lower than expected, indicating that the chemical cleaning regimes were not always totally effective. Increasing the solution cleaning temperature (to 30° C, mainly during the winter) and occasionally performing a citric acid CIP after the primary cleaning regime may help to improve the cleaning efficiency.Generally throughout the test period the quality of the reverse osmosis permeate was very good, with conductivity values ranging between 22 and 35 microsiemens/cm. Normalized salt rejection for the Ramos Arizpe pilot tests averaged 99.7%. The average normalized and non-normalized (actual operating value) permeate fluxes for the Ramos Arizpe pilot tests averaged were 21.4 and 9.3 gfd versus average fluxes of 8.3 and 8.4 gfd for the Silao unit. For the Ramos Arizpe pilot unit, the non-normalized values are lower due mainly to operating at low feed pressures (in comparison with the factory test pressure) in order to reduce membrane fouling. The Silao units operated at feed pressures closer to the factory test pressure and therefore, the normalized and normalized flux values are nearly the same. The average Silt Density Index (SDI) for the RO feed measured during the pilot test at Ramos Arizpe was 1.5. The maximum SDI was 4.2. Two of the membranes used at Ramos Arizpe were cleaned only once at the end of the test, at a specialty supplier's facility using three of their products. Flux recovered to the membrane nominal value. The membrane size and array configuration limited the recoveries at both RAC and Silao to 60%. For the full-scale system, it is expected to operate at recoveries on the order of 70%.Based on the pilot test a full-scale system for re-using treated effluent is technically feasible. An improved pretreatment however, in terms of phosphate and COD removal will allow increased recovery, reduce membrane fouling, and reduce the frequency of chemical cleaning.