Pilot Testing Methodology: How to Design a Membrane Pilot That De-Risks Capital
A membrane pilot program is one of the most valuable investments a project team can make before committing full capital to a water treatment system. Done well, it eliminates the technical unknowns that cause budget overruns and performance disputes during commissioning. Done poorly, it generates ambiguous data that delays decisions without reducing risk.
Define the Decision the Pilot Must Support
Before designing any aspect of the pilot, define the capital decision it must enable. Is the goal to confirm membrane technology selection and size a full-scale system? To validate pretreatment adequacy for a proposed RO installation? To demonstrate performance to a regulator or project financier? The answer determines every design choice that follows — scale, duration, instrumentation, and what constitutes a “successful” pilot.
Feedwater Characterization First
Comprehensive feedwater characterization before pilot design is non-negotiable. Sampling at multiple points in the water cycle, across seasons, and under different operating conditions provides the input data for scaling analysis, membrane selection, and pretreatment design. Surprises during a pilot almost always trace back to incomplete or non-representative pre-pilot water sampling.
GWTS routinely performs full water chemistry panels including: TDS, TSS, TOC, TIC, pH, turbidity, conductivity; full ion panel (Na, K, Ca, Mg, Ba, Sr, Cl, SO₄, HCO₃, CO₃, SiO₂, B); metals by ICP-MS; oil & grease (EPA 1664); and SDI/MFI per ASTM D4189.
Scale and Duration
Pilot scale should be large enough to produce statistically meaningful flux and rejection data. For MF/RO systems, bench-scale testing (0.1–1 GPM) is appropriate for membrane screening and pretreatment optimization; pilot-scale testing (1–10 GPM) provides the flux, fouling, and CIP data needed to design a full-scale system.
Duration is often underestimated. Short pilots (1–2 weeks) can demonstrate initial performance but miss seasonal feedwater variation and the fouling trajectory that matters most for lifecycle cost projection. GWTS recommends a minimum of 4–8 weeks for a meaningful RO pilot on produced water, with longer programs (3–6 months) where seasonal variation is a known factor.
The Deliverable: A Bankable Basis of Design
A pilot program is only as valuable as its final report. The deliverable should be a formal Basis of Design document that translates pilot data into full-scale system design parameters: design flux, recovery, element selection, pretreatment specification, CIP protocol and frequency, and projected operating costs. GWTS pilot programs deliver complete performance data records, fouling and CIP analysis, full-scale system design basis, and capital and operating cost estimates for the recommended full-scale configuration.