Make a weapon to destroy COVID-19.

What it does

Irradiates air dosing at 2.0mJ/cm2 of UVC, which scrambles the DNA of the virus, rendering it inactive@>95%. Even this small unit has airflow @ 4.5m3/min, units could be deployed to hospital waiting rooms, treatment rooms, triage centers etc. We can greatly reduce the risk of small airborne particle infection for medical staff.

How I built it

Household fan, 6 X 6 watt UVC units, aluminum duct, tape, scraps from my workshop. If these parts are not already in hospitals they can be easily acquired.

Challenges I ran into

None, very easy to put together

Accomplishments that I'm proud of

The unit can be built from parts already in-hospital or available online and/or home centers. Hospitals have UVC units to sterilize instruments, they could be re-purposed into Corona Killers.

What I learned

The gap between 95% inactivation and 99.999% is huge in terms of energy and critical components required, it is quite the chase down the diminishing marginal returns curve. With the same number of UVC units we can make 3 high air volume units with 95% inactivation or 1 hopelessly low air volume unit with 99.999% efficacy. I know that doctors and health officials expect five nines sterility but for treatment rooms filled with corona patients I suggest that we need huge volumes of pretty good air instead of tiny volumes of sterile air.

The calculations dimensions, intensity, and max airspeed need to done by someone with science or engineering background or the unit will not be effective. Radiologist, X-ray Tech etc could easily run the calcs based on the dimensions of parts already in the facility.

What's next for Corona Killer 1.0

Attack the virus at close quarters. Ventilators are a point of infection for medical staff. The combination of a patient in respiratory distress, bodily fluids, pressurized air, this is number one place we must inactivate the virus before it reaches medical staff. We want to build a Far UVC (220nm) unit to protect health workers. Far UVC does not penetrate skin or eye membranes so can be used to locally irradiate high-risk locations in medical or retail(grocery stores, pharmacies).

Think not the last mile problem, think the last 50cm problem. I want to create a zone of >95% right in front of medical staff who are treating patients. We are thinking about a headlamp style or chest-mounted. Power to come from battery pack held on belt. 6-12watts from far-UVC LED array would render the virus inactive in less than a second. Assuming they are wearing some sort of face shield to protect from a direct attack such as a cough, any airborne particles would get dosed before they reach the nose/mouth/eyes of the medical worker.

Details and calculations of Corona Killer 1.0, large room/large space COVID-19 killer.

We believe UVC technology is a tool available to us, but has not been explored.


Right now we need more tools to slow progression, to flatten the curve. I believe UVC technology is one more tool available to us. The key is to do it in a way that is quickly deployable and effective.


A pipe or rectangular box open on both ends, loaded with UV LEDs of the appropriate count and wattage. Air is pushed through by fans, blowers already on site, the result is an air supply >95% virus free.


Reports say that COVID-19 can be viable in air for up to 3 hours. UV light, in particular UVC has fantastic anti-viral, anti-microbial efficacy. That is too long and will lead to secondary infection, especially in high-traffic areas. We need to deploy units in grocery stores, hospitals etc where people must gather.


UV light, in particular UVC has fantastic anti-viral, anti-microbial efficacy. Using the specifications outlined below, we believe these units are both cost effective and can be easily built.


The starting point is a dosage number, the amount of energy required for virus inactivation. This article suggests the dose of 2mJ/cm2 is sufficient for 95% lethality/inactivity using UVC.

In my calculations below I am assuming the target dosage and UVC wavelength as suggested in the article.

In order to make the calculations, here are the variables. Pipe with radius r and length l UVC LED array with wattage W, field of transmission in %, length of LED array determines l I have made a number of assumptions, that any engineer/specialist will be quick to point out. I am aware of these assumptions and believe my calculations reflect the cautious side of the assumptions. An engineer on site with parts in his hands will be able to adjust the formula to make more accurate calculations. It is my intent to have grossly calculated on the side of caution.

From parts available online and at home centers 36W, 44cm UVC LED array, l is 44cm, 10cm r pipe Any pipe is fine if not aluminum then line with aluminum tape, shiny aluminum will increase internal reflection. What we need to calculate is how much time the air must spend in the pipe to achieve the target dosage.

Let’s convert the wattage to milliwatts 36W *1000mW=36,000mW Given the size of pipe we can calculate its surface area Surface area of cylinder= A=2πrh+2πr2 A=3399cm2 The UVC lamps will not strike the entire inside surface of the pipe, it has limited field of transmission. This will need to be adjusted for each light source. Via measurement 1.6% of the internal surface will not be irradiated by the UVC lamps so will subtract from total surface area. 100%-1.6%=98.4% Total surface area 3399cm2 X 98.4%=3345cm2

So then we calculate how much time it will take to give us target dosage. wattage/effective surface area 36,000mW /3345cm2=10.76mW/cm2 intensity

1 milliwatt equals 1 millijoule per second. 10.76mW/cm2 X 1mJ/s=10.76mJ/s/cm2

target energy/intensity=transit time 2mJ/cm2 / 10.76mJ/s/cm2=0.186 seconds max transit time

And given the l of the array, the max speed to send the airflow. l=v/t or l/t=v .441m/.186 seconds= 2.37m/s max airspeed for >95% efficacy. This airflow could be provided by a standard fan at low or med speed. Video shows the fan to be at approx. 1.9m/s, well within the max airspeed.

FAQ -is it scaleable? Yes it is. Pretty sure I saw 1m diam fans/blowers at both SEATAC and SFO. That would be a good place to start. At that size: 240W UVC LED, r=50cm, l=300cm, max airspeed 4.4m/s Yes, the unit would be 3 meters long, but it would process 204m3 of air per minute. A couple of these units could cover an entire arrivals area.

-safety? Need to attach a cover to intake and exhaust.

Built With

Share this project: