The Second Site Is a Different Business
A single hydroponic farm is managed by attention. The owner or head grower walks the crop, notices a slow irrigation valve, hears a bearing starting to fail, and fixes it. That model works well and scales not at all.
A multi-site hydroponic farm operation has to replace attention with standards, measurement and reporting. Sites that look identical on a drawing behave differently in reality: different water, different local labour, different climate, different market. The management problem is to keep the parts that should be identical genuinely identical, while leaving each site free to handle what is actually local.
What Should Be Standard Everywhere
| Category | Standardise | Allow local variation |
|---|---|---|
| System type and components | Yes — one design, one spares list | Capacity and bay count |
| Nutrient recipes | Base recipes and target ranges | Adjustments for local water analysis |
| Crop varieties | Preferred list with regional substitutions approved centrally | Varieties chosen for local market |
| Operating procedures | Every safety-critical and quality-critical task | Shift timings and local work rules |
| Data and reporting | Metric definitions, units, reporting cadence | Nothing — this is where variation is most expensive |
| Procurement | Preferred suppliers and part numbers | Local consumables where quality is proven |
| Training | Core curriculum and certification | Language and delivery method |
The test for whether something should be standardised: does variation here prevent me from comparing performance across sites, or create a safety or food-safety risk? If yes, standardise it and keep it written down.
Reporting: Same Numbers, Same Definitions
Cross-site reporting fails for one reason almost every time — the same word means different things at different sites. “Yield per square metre” is meaningless if one site measures growing area and another measures total floor area. Fix the definitions before you fix the dashboard.
- Fix the denominators. Growing area, net saleable area, and packhouse area are three different numbers; pick one and use it everywhere
- Fix the period. Weekly reporting on a Monday-to-Sunday basis compares properly; a rolling period does not
- Fix what counts as a defect. One site’s “seconds” is another site’s “waste”, and the gap shows up as a fictional performance difference
- Include leading indicators, not just yield: irrigation EC and pH drift, days since sensor calibration, filter differential pressure, labour hours per crate
- Report exceptions, not everything. A weekly exception report — this is outside range and here is what we did — is read; a 40-page dashboard is not
- Close the loop. Every exception should have an owner and a date, and the following week’s report should show whether it was fixed
Remote Monitoring: What It Is Good For
Remote monitoring earns its place in a multi-site operation, but not by replacing people. Its job is to tell you where to send attention.
| What to monitor remotely | Typical alarm value |
|---|---|
| Irrigation pump running status and pressure | No flow during a scheduled cycle |
| Reservoir level and EC/pH | Outside the crop’s range for more than a set interval |
| Room and root-zone temperature | Outside band, or rate of change too fast |
| Chiller and heater status | Running continuously, or not running when it should |
| Power supply and backup | Mains failure, generator start, battery voltage |
| Door and access state | Out of hours, which usually signals a procedure failure |
| Water consumption per day | Sudden increase, which usually means a leak |
Two disciplines make remote monitoring useful rather than noisy: alarm thresholds must be set per site from that site’s own baseline, and every alarm must have a defined response. An alarm nobody acts on teaches the whole team to ignore alarms.
Staffing Across Sites
Three models appear in practice, and most operations end up with a hybrid.
- Autonomous sites with a competent site grower. Fast decisions, high local ownership, risk of divergence and of the site grower becoming a single point of failure.
- Central technical function with local operators. Strong consistency and shared spares, but central staff become a bottleneck and local initiative can fade.
- Hub and spoke. One large site hosts technical, agronomy and maintenance functions that serve smaller satellites. Usually the best balance once three or more sites are running.
Whichever model you pick, a shared technical standard and a rotation of staff between sites are the two cheapest consistency tools available. People who have worked at two sites implement standards better than people who have read them.
Failure Modes That Appear Only at Scale
- Spares drift. One site stocks a different pump model; suddenly you hold four part numbers for the same function
- Procedures diverge quietly. Each site makes small local edits until the “standard” exists only in the manual
- Reporting becomes theatre. Numbers are produced because they are requested, not because anyone acts on them
- Central control loses credibility after a decision made without local knowledge causes a crop problem
- Key-person dependency at each site, invisible until someone takes leave
- Capital decisions repeated. Each new site re-tenders everything and loses the benefits of the first purchase
All six are management problems rather than technology problems, and all six are cheaper to prevent with written standards and a shared metric set than to fix later.
FAQ
How many sites can one manager oversee?
Fewer than people expect if each site is genuinely independent. With good remote monitoring, standard procedures and a strong site grower at each location, one technical manager can support several sites — provided the reporting is honest and comparable.
Should all sites use the same system design?
Yes where possible. Standard designs reduce spares, training and troubleshooting time. Deviate only for a documented reason such as water quality, climate or building constraints.
What software do I need?
Enough to record batch data, exceptions and maintenance across sites in one system. That can be a purpose-built platform or a disciplined spreadsheet process — what matters is that definitions are identical and the data is actually reviewed.
How do I compare sites fairly?
Compare like with like: same metric definitions, same reporting period, and normalised for climate and market price. Otherwise a site with a cheap local market or an unusually hot summer looks badly managed when it is not.
How do I keep standards from drifting?
Audit against them. A short internal audit twice a year, scoring each site against the standard, is usually enough to keep divergence small — and it doubles as preparation for external certification audits.
Is remote monitoring safe on a public network?
Treat it as an industrial control system: separate the control network from office IT, change default credentials, restrict remote access, and keep physical control available if the network fails. A farm that cannot be operated without the internet is a farm with a single point of failure.
Standardise Before You Duplicate
If you are planning a second or third site, the most valuable work happens before construction: fixing the design, the metric definitions and the procedures so that the next site starts from a standard rather than from scratch. Send us your site plans through the quote form and we will help build the standard.
Related reading: standards depend on measurement — see smart hydroponics and remote monitoring and traceability and batch records, and reuse one design with OEM and private label manufacturing.