With no plant yet built and no measured data to draw on, the engineer has to substitute analogous, calculated and pilot-derived information for the direct measurement an in-plant survey would otherwise provide, and then build in conservatism/flexibility to cover the resulting uncertainty:
Literature and published-data review — search for waste generation factors from any published effluent guideline, industry-association data, or case study for the same or an analogous process/SIC code (an unbuilt "unicorn" process rarely has a direct precedent, so the closest analogous manufacturing process — e.g. similar raw materials, similar unit operations, similar production scale — is used as a starting benchmark).
Mass/material balance from the process design — use the proposed process flow diagram, raw-material inputs, product yield and process chemistry (stoichiometry) to calculate waste generation directly: waste = raw material input − product output − recycled/recovered material, for both liquid and solid streams.
Equipment vendor/technology-supplier data — the manufacturers of the specific process equipment being purchased can typically supply water-usage and waste-generation rates for their own units, since they have tested them.
Bench-scale or pilot-scale testing — run a scaled-down version of the actual proposed process (even a laboratory bench-scale reactor) to directly generate and characterize representative waste samples before full-scale construction.
Site visits/benchmarking against comparable existing facilities — if any comparable operating facility exists elsewhere (even in another jurisdiction), review its published discharge monitoring reports or arrange a site visit to obtain real operating data.
Regulatory-agency and permit-database review — publicly available discharge monitoring reports from similarly-permitted facilities can substitute for direct measurement.
Conservative engineering estimates with a design contingency — where no better data source exists, apply a documented, defensible engineering estimate with a safety factor, rather than an unsupported guess.
Phased, flexible/expandable design — because pre-construction estimates always carry uncertainty, design the treatment system with built-in flexibility (modular capacity, monitoring provisions) so it can be adjusted once actual production data becomes available after start-up, rather than betting the whole design on an untested estimate.
