Algae cultivation: a light sensor built to live in seawater

The situation

A marine research group cultivates algae in outdoor seawater tanks. The work needs the growing conditions in each tank on the record, and for algae the condition that matters most is light, because light is the input photosynthesis runs on. Dissolved oxygen and pH follow, both of them responses to how the culture is doing.

Seawater is the complication. It is corrosive, it is conductive, and it works its way into any joint that was designed for fresh water and hoped for the best. Probes that are reliable in a freshwater tank are not automatically reliable here, and one measurement in particular had nowhere to go.

The gap in the market

There was no shortage of light sensors. There was a shortage of light sensors that could sit in seawater indefinitely. A light sensor has to have an optically clear window, and that window is a sealing joint positioned exactly where the sensor is most vulnerable. Dry-land sensors put a cheap gasket there. Immersion changes the requirement completely, and continuous immersion in seawater changes it again.

So we built one. The sensor reads 0 to 20,000 lux, and it was designed around the housing and the cable entry rather than having them added at the end, because those are the parts that decide whether a submerged instrument lasts a season or a fortnight. It speaks the same protocol as every other sensor in the range, so from the controller’s point of view it is simply another instrument on the bus.

What was installed

Each tank has an Omni Genesis controller reading three things from the water below it: light from the custom sensor, dissolved oxygen from a DO-110 titanium probe, and pH from a PH-110 saltwater probe. The titanium body on the DO-110 and the saline specification on the PH-110 are the reason those particular variants are in the catalog; the standard versions belong in fresh water.

Powering a row of controllers outdoors was its own problem. A wall adapter per controller is a collection of failure points in the weather, so we built a cabinet instead. One mains feed goes in. Protective breakers and a regulated supply sit at the top of the enclosure, then an individually fused output channel per controller, each leaving through its own gland. It carries up to 16 controllers, and a fault on one tank does not take the row down with it.

Why it matters

Two things in this deployment did not exist as products before the project needed them: a light sensor that survives seawater, and a way to power a row of controllers cleanly outdoors. Neither required a new platform. The sensor joined the same bus as everything else and the cabinet feeds standard controllers, so the custom work stayed at the edges where the requirement actually was.

That is what modular hardware buys a research group. Closed systems can offer whatever their vendor decided to build. An open one can be extended to measure the thing your work actually depends on, even when nobody sells it yet.