A Shot at the Spark

The first ten minutes

Every great bushfire was once small enough to put out with a bucket.

Most of the land that burns in a bad Australian season is burned by fires that lightning started. In Black Summer, 2019–20, lightning fires accounted for more than 90 per cent of the area burned. A dry thunderstorm can start dozens of fires in one evening, in steep, remote forest with no road and nobody nearby.

In its first ten to thirty minutes, a new ignition is a few square metres of burning litter. It is easy to stop. By the time a crew walks in, or an aircraft is free, hours or days have passed and it cannot be stopped at all.

The problem is not putting the fire out. It is the delay. On extreme wind days, faults on power lines start fires the same way, often far from help.

So consider a network of stations that see the strike, confirm the fire, and put foam on it within minutes, then keep it small until crews arrive, without anyone having to get there first.

The idea

Three parts, each covering another’s weakness.

• Rounds, for speed. Each station keeps a magazine of gliding rounds. When a new fire is confirmed, it fires a volley. The rounds find the flames by their heat and burst foam just above them, about eight minutes after the strike.
• Tankers, for persistence. A swarm of small uncrewed planes follows the rounds out. They fill up at the nearest dam and keep pouring water on the fire, in turns, until crews arrive.
• Launchers that move, for reach. Fixed stations hold the best hilltops. When a dry storm is forecast, mobile launchers drive out and fill the gaps in its path.

Rounds are fast but spent. Tankers are slower but come back, again and again. Mobile launchers put rounds where the storm is going, not just where the stations happen to be.

How a fire is found

Most of the detection already exists or is being built.

• Lightning networks. National ground networks already record where lightning strikes. Research in Australia is now picking out the few strikes most likely to start a fire: those with a long continuing current, which heats the fuel for longer.
• Satellites. FireSat, a constellation built to spot fires as small as about 5 by 5 metres, launched its first satellites this year. Early on, it will pass over each region only a few times a day.
• The station’s own eye. A thermal camera on each station’s mast watches its own country, so a flagged strike can be checked at once.

The gap is the last step: confirming that a particular strike has started a fire, and reaching it before it grows.

The round

• Thrown by compressed air. The launcher uses air, not an explosive charge, so there is no spark in dry fuel. It is quieter, simpler to approve, and cheap to reload.
• Fired steeply, then glides. Near the top of its climb, the round swings its wing open and glides to the fire, 10 to 20 kilometres away.
• Finds the fire by its heat. The lightning network and the camera give a rough position. An infrared camera in the round’s nose locks onto the hottest point and steers to it.
• Bursts above, not on. Just above the fire, a gas cartridge drives the foam out in a cloud that settles over the flames.
• Harmless if it misses. It carries no explosive, and its casing breaks down in the soil.

One round may not be enough. A station fires three to five at each new fire.

The tankers

The rounds knock a fire down. Something has to keep it down, sometimes for hours, until a crew can walk in.

• Small planes, no pilots. A swarm of uncrewed tankers, each carrying a few hundred litres. Small enough to be cheap and to fly low under the smoke, large enough to matter in relays.
• Water from the nearest dam. Farm dams and reservoirs are scattered through the bush. A tanker fills up at the nearest one, flies a few kilometres, pours, and goes back. A little foam concentrate in the tank makes the water cling to fuel instead of running off.
• One always over the fire. The swarm takes turns, so the fire is never left alone for long. The more tankers, the shorter the gap.
• Eyes for the crews. Each tanker carries a thermal camera. Together they show where the fire is, whether it is out, and where it is going, and they keep watching the slopes after a storm for fires still smouldering.

They are based at the larger stations, launched from a rail and landed on a short strip, so nobody needs to fly them out from a city airport.

Stations that move

Fixed stations cannot be everywhere. But dry lightning storms and extreme wind days are usually forecast a day or more ahead, so the launchers can move first.

• Into the storm’s path. The evening before a forecast storm, mobile launchers drive out along fire trails and park in the gaps between fixed stations.
• On prepared pads. Each one fires from a cleared, surveyed pad beside a trail, so it is ready the moment it arrives, and nothing around it can catch fire.
• More rounds where they are needed. A storm that starts many fires can empty a station’s magazine. A mobile launcher brings its own, and can restock the fixed stations nearby.
• Along power lines on wind days. On extreme wind days, they can wait beside the lines most likely to fail.

Most would be all-terrain trucks, which are faster on roads. Tracked launchers are for steep country off the trails. Either way, their value is in being in place before the first strike. A vehicle that sets out after the fire starts is an hour or more away, which is exactly the delay this design is meant to remove.

Fewer fixed stations are needed if mobile ones cover the gaps on the days that matter.

Borrowed from war

Almost every part of this has already been built, and most of it was built for war.

• Drones that leave a tube with their wings folded and open them in flight are now made by the thousand.
• Infrared seekers that find the hottest point in a scene are decades old.
• Drones that wait over an area and act only when something appears are a whole class of weapon.
• Launchers on tracked vehicles that drive in, fire and move again have been standard for decades.
• Software that flies many drones together, and the loop that locates where a shot came from and answers within minutes, are standard military engineering.

Wars have spent enormous sums making these parts cheap, reliable and fast. The engineering does not care what it is pointed at. I would much rather it were pointed at the first flame of a fire than at a person.

The civilian version should be civilian all the way down: commercial thermal cameras and autopilots, foam instead of explosives, and every station run openly by a fire agency.

A sketch of the numbers

Rough figures, to show the scale.

Quantity	Value
Foam solution per round	≈ 10–40 kg
Foam from 20 kg of solution, expanded ten times	≈ 200 litres
Area it covers 2 cm deep	≈ 10 m²
Glide range	≈ 10–20 km
Height needed to glide 15 km at a glide ratio of 10	≈ 1.5 km
Flight time for 15 km at about 50 m/s	≈ 5 minutes
Area one station covers, 10–20 km radius	≈ 300–1,250 km²
Public forests and parks in Victoria	≈ 80,000 km²
Stations to cover them	≈ 60–250
Rounds per fire	3–5
Water per tanker	a guess: ≈ 200–500 litres
Round trip for a tanker, dam 5 km from the fire, at 30 m/s	≈ 6 minutes plus refilling
Warning of a dry lightning day	often a day or more
Distance a mobile launcher covers in 3 hours, roads and trails	≈ 100 km
Cost per station and per round	not yet known

The numbers that matter most are not in the table yet: how much foam it takes to stop a small fire in dry fuel and wind, and what one Black Summer would have cost against a few hundred stations and their tankers.

The hard parts

Enough foam. A round’s load may be enough to smother a fire of a few square metres in still air. It is not known yet how much it takes in wind, on a slope, in thick dry fuel. This is the first experiment.

Getting high enough. To glide 15 kilometres, a round must climb to about 1.5 kilometres. Compressed air alone may not throw it that high. It may need a small electric motor for the climb.

Accuracy in wind. Gusts and the fire’s own updraft will push the round around in the last few seconds. The foam cloud has to be large enough to forgive a miss of a few metres.

Ridges. In steep country, a ridge hides the valley behind it from the station’s camera. Stations need to be placed so that they overlap, mobile launchers can fill the worst gaps, and satellites can see what the cameras cannot.

Many fires at once. One dry storm can start more fires than a station has rounds. Stations would need to share targets, mobile launchers would need to restock them, and the agency would need to decide which fires to hit first.

Flying the tankers. Low over ridges, in smoke, in gusty wind, in airspace shared with crewed water bombers. They need to see and avoid each other and any aircraft, and land safely when a dam is empty or a tanker fails.

Moving the launchers. Fire trails are slow and rough, and tracks churn them up. A hot exhaust can start a fire in dry grass. A crewed vehicle in the bush on a lightning day puts its crew near the fires, so the crew should park on the pad and withdraw, leaving the launcher to be fired remotely.

People in the bush. Bushwalkers and campers may be near a new fire. A round must never be fired where it could hit someone, and the airspace has to be cleared with CASA and with any firefighting aircraft already flying.

Remote stations. A station on a remote hilltop must survive heat, storms, animals and theft, and be serviced rarely. It is also in the path of the fires it is there to stop.

How it looks. “Firing rounds into the bush” sounds alarming. The design has to be explained plainly: no explosives, foam that breaks down, and a person at the agency approving every volley.

Cost. Hundreds of stations are expensive. So is one bad season. The comparison has not been done.

Where I would start

• On paper first. Glide-range and response-time models, and a map of where stations would go in one Victorian forest region, checked against where lightning fires have started in the past.
• Foam trials. Controlled test burns, to learn how much foam stops a spot fire in real fuel and wind.
• One station. A prototype on a single hilltop, firing at test fires under the supervision of a fire agency, with one mobile launcher and a few tankers beside it.
• A prior-art search. Each part exists separately. The combination appears new, and a provisional patent may be worth filing.

Others who have tried parts of this

• Firefighting drones already carry water and foam to small fires and to tall buildings.
• Uncrewed water bombers. Helicopters and planes flown without a pilot have been trialled for water bombing.
• Fire-extinguishing projectiles and throwable extinguisher balls have been built and sold.
• Tube-launched drones with folding wings are a mature design.
• Lightning-risk detection research, in Australia and elsewhere, is learning which strikes are likely to start a fire.
• FireSat and other satellites are being built to see small fires within minutes of their start.

What I add is the combination: stations that see the strike, rounds that find the flame by its heat, launchers that move into the storm’s path, and tankers that stay with the fire until people arrive.

The best time to fight a bushfire is the first ten minutes. Nobody has been there for them, until now.