At my workplace we are warned against using compressed air to blow dust of working clothes etc, except when a special nozzle is affixed to the to the hose, reducing the pressure of the released air.

One danger that was mentioned at a workshop I was attending is that if the air is directed at your skin, some air may penetrate the skin and create air bubbles in your blood which will cause an embolism.

A quick internet search brings up the same claim on e.g. the page Engineered Systems, which states:

Compressed air blown into the skin can obstruct an artery and result in an embolism. This is a condition where a pocket is created by the blast of air inside a blood vessel. Once this pocket of air enters the brain or heart, it can lead to stroke or sudden cardiac arrest.

I see similar claims on several other sites as well, but just the claim nothing backing it up or any actual cases of this ever happening.

Is it accurate that pressurized air directed on the skin can cause an embolism? The system pressure in our facility will typically be between 6 and 7 bars, but I'd also be interested in knowing if it can happen at higher or lower pressures. Are there any documented cases of this having happened?

  • 2
    The first google hit I get on this says "When high-pressure compressed air is used to clean skin and clothing, it can penetrate the skin and enter the bloodstream." I have no doubt that if it does enter the blood it causes an embolism... by definition. I suppose the real Q is how likely that is to happen by blowing compressed air onto the skin. Aug 16, 2022 at 11:21
  • 4
    Several articles (example) mention that the threshhold for skin penetration is 100 psi (~6.8 bar). That's significantly lower than I expected... Aug 16, 2022 at 11:27
  • 2
    If the threshold for skin penetration is 100 psi, would that imply that the nozzle is pressed against the skin? What does 100 psi at a nozzle two feet from the skin create for pressure on the skin?
    – fred_dot_u
    Aug 16, 2022 at 15:42
  • 2
    @fred_dot_u 100 psi at 2 feet would be closer to 0 psi
    – jesse_b
    Aug 16, 2022 at 15:55
  • 2
    @jesse_b, yeah, distance is probably a critical factor here, which is why this is not easy to quantify. Contains NSFW pictures sciencedirect.com/science/article/pii/S2090536X16300508 Aug 16, 2022 at 19:05

1 Answer 1


To answer your specific questions,

Is it accurate that pressurized air directed on the skin can cause an embolism? ... Are there any documented cases of this having happened?

Yes, there is at least one documented case. I found it in the OSHA (US Occupational Safety and Health Administration) industrial accident database using the keyword "embolism" and scanning through the data for any mention of compressed air. The accident summary states:

At 8:30 a.m. on May 20, 2019, an employee was using a compressed air gun to clear a paper jam from a Gambini Rewinder Machine. The employee lost his grip on the air gun and cut his hand. The laceration to the employee's hand caused an embolism to form requiring hospitalization.

This is the only case I found, so it appears to be fairly rare.

OSHA Nozzle Requirements

The "special nozzle" that your workplace requires is described in OSHA standard 1910.242(b):

Compressed air used for cleaning. Compressed air shall not be used for cleaning purposes except where reduced to less than 30 p.s.i. and then only with effective chip guarding and personal protective equipment.

This is further clarified in OSHA Instruction STD 01-13-001 - STD 1-13.1 October 30, 1978 which states:

The phrase "reduce to less than 30 psi" means that the downstream pressure of the air at the nozzle (nozzle pressure) or opening of a gun, pipe, cleaning lance, etc., used for cleaning purposes will remain at a pressure level below 30 psi for all static conditions. The requirements for dynamic flow are such that in the case when dead ending occurs a static pressure at the main orifice shall not exceed 30 psi. This requirement is necessary in order to prevent a back pressure buildup in case the nozzle is obstructed or dead ended. See enclosure (1) for two acceptable methods of meeting this requirement. Also, there is no intent to restrict the diameter of the nozzle orifice or the volume (CFM) flowing from it.

It appears that the standard is intended to ensure that the pressure on the skin when the nozzle is fully obstructed by a body part is no greater than 30 psi. The standard does not prohibit higher pressures in unobstructed nozzles.

Embolism Risk

While there are a number of sources that mention the risk of embolism, I didn't find that in any OSHA documents. I did see it in an Oregon OSHA fact sheet which says that compressed air can enter the bloodstream through an existing wound.

Compressed air can enter the body through a scratch or small puncture wound and cause swelling and severe pain. If the air gets into the bloodstream, it can cause an air embolism, which can cause symptoms similar to a heart attack or a stroke.

In the OSHA accident report cited above, the embolism occurred as a result of the employee lacerating his hand, and not by compressed air entering through the skin.

I spent some time looking for a basis for the OSHA 30 psi requirement, but was not able to find anything. Perhaps someone else will be more successful.

  • 5
    I think it's important to note that in the documented accident, the worker cut himself, exposing the blood vessels so that the compressed air could enter. It does not say anything about compressed air penetrating the skin and entering the blood vessels this way. Aug 19, 2022 at 10:05
  • @SebastianRedl - good point, I added it. Thanks.
    – Mark
    Aug 19, 2022 at 16:51
  • @DanilaSmirnov linked an article in a comment on the OP which seemed well sourced. In particular references 6 and 7 look like they may provide cases of pressurized air entering through the skin, while reference 8 may provide some basis for the OSHA requirement. Unfortunately I don't have access to the full text of the references, so I can't provide much help besides linking the article. Anyway, I guess this answer does answer the question to a certain extent, so unless a better one pops up in the next few days I'll give you the checkmark
    – eirikdaude
    Aug 24, 2022 at 9:25

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .