RAS System Buyer's Guide & Tender Preparation

1. Frame the project before you talk to suppliers

Many RAS problems start before design — the moment a sponsor asks a vendor 'how much for 500 tonnes?' without a written brief. Vendors will happily quote, but the numbers are meaningless until you fix the biology, the species life-stage, the site water and the energy envelope. Document these before any RFI.

• Species, life-stage, target size at harvest
• Annual production (t/year) and ramp-up curve
• Source water analysis (all major ions, temperature)
• Grid capacity (MW), tariff and reliability
• Discharge permit envelope
• Target unit economics ($/kg)

2. Design the process, not the equipment

A RAS is a mass-balance problem. Feed in → oxygen consumed, CO₂ produced, TAN produced, sludge produced. The equipment list is the consequence of that balance. Ask suppliers for their mass-balance calculation, not just a P&ID. Any vendor that won't share the calculation is not the right partner for a bankable project.

• Design feeding rate (kg feed / day at peak biomass)
• O₂ demand (typically 350–450 g O₂ / kg feed)
• TAN production (typically 25–35 g TAN / kg feed)
• CO₂ production (typically 1.375 × O₂ consumed)
• Sludge production (~30% of feed dry matter)
• Make-up water (typically 300–1,500 L/kg produced)

3. Specify performance, not brands

Employer's Requirements should describe outcomes: dissolved oxygen ≥ 6 mg/L, TAN ≤ 1 mg/L, CO₂ ≤ 15 mg/L, availability ≥ 98.5%, specific energy ≤ agreed kWh/kg. Do not name pumps, drum filter models or biofilter media. Naming brands hands negotiating leverage to the vendor and locks you out of a competitive tender.

4. Understand the CAPEX bands

As a rough planning range only (not a quote), turnkey RAS CAPEX for salmon-class projects is often discussed in a $12,000–$25,000 per annual tonne band once civil, MEP and biology working capital are included. Warm-water species (tilapia, barramundi, shrimp) sit lower ($6,000–$14,000/t). These ranges vary widely by site, scale and scope; a quote outside them is a reason to ask for the mass-balance and a scope schedule, not proof that it is wrong.

5. OPEX is where projects live or die

Energy and feed are usually the largest OPEX lines for a RAS grow-out. Insist on 10-year OPEX modeling with an energy escalation scenario. Ask every bidder to state the eFCR their design assumes and the conditions behind it; feed conversion depends on species, feed, temperature and management, so treat any figure as an assumption to verify, not a guarantee.

Comparing proposals line by line

Put every proposal into the same table before looking at price. Missing information should stay visible as 'not stated' — never fill a gap with your own guess.

• Species, stages and harvest size each design assumes
• Rated production capacity and the stocking density behind it
• Water source, make-up water and discharge assumptions
• Filtration, biofilter, degassing and oxygenation scope
• Backup power, backup oxygen and alarms
• Monitoring, controls and data capture
• Exclusions: civil works, buildings, permits, utilities
• Installation, commissioning, training and acceptance tests
• Service coverage, spare parts and response times

6. Run a two-envelope tender

Technical envelope opens first. Only bidders that pass technical compliance move to commercial. This is standard for banks and DFIs and protects the project from a race-to-the-bottom on price.

7. Structure warranty and LDs correctly

24-month mechanical warranty from Commercial Operation Date. Performance liquidated damages tied to guaranteed capacity, eFCR and energy. LD cap of 10–15% of contract value. Total liability cap at 100%. Insist on parent-company guarantee where the local entity is thin.

Size it before you tender

FAQ