The amount of dry ice a blasting job needs cannot be read from one industry average. A defensible first order starts with the selected machine feed rate and estimated trigger-on time, then adds expected storage and handling loss plus a stated contingency. After the trial, replace that estimate with measured dry ice per accepted part, area or cleaning cycle.
That distinction matters. Feed capacity describes what a machine can meter. It does not describe how long the trigger will be on, how quickly the process reaches the acceptance standard, or how much dry ice will sublimate before use.
Start with the consumption equation
Use one transparent purchasing equation:
Initial purchase mass = selected feed rate × planned trigger-on time + expected pre-use loss + stated contingency
Keep every term in the same unit. Do not multiply the feed setting by the full shift unless the machine will actually blast for the full shift. Setup, inspection, repositioning, refilling and production access reduce trigger-on time even when the crew remains assigned to the job.
For recurring work, the stable measure is different:
Media intensity = dry ice actually consumed ÷ accepted cleaning output
Accepted output may be a part, a mold cavity, a square metre or a completed maintenance cycle. The denominator must include only work that passed the documented cleanliness and surface criteria. Otherwise a high cleaning rate can hide rework or rejected parts.
Published rates are boundaries, not forecasts
Current supplier figures show why a single answer is unreliable. Cold Jet’s FAQ gives an average equipment range from approximately 0.32 kg per minute for a MicroParticle system to 1.1 kg per minute for a pellet system. ICS lists dry ice consumption to 24 kg/h for its compact IC 022 and a 0–110 kg/h setting range for the higher-output IC 410.
These numbers describe different machines and configurations. They do not prove that one system cleans faster or uses less dry ice per accepted part. A low maximum feed can suit delicate or small-area work but limit removal capacity on a heavy deposit. A high available feed rate can support demanding work, yet it does not mean the operator should use the maximum.
The live search results also included a supplier guide quoting a much wider average-usage range without tying it to a named machine. That is useful counterevidence: an unqualified web average is not a procurement specification. Use the exact model’s data sheet to set the possible range, then use a representative trial to find the working point.
Four variables control the real order
Feed setting and machine architecture
The machine determines whether dry ice feed is fixed, stepped or continuously adjustable. Cryoblaster’s current FAQ notes that some machines are rated to a set consumption, while adjustable feed can be matched to the contaminant and pressure. For a buyer, the relevant questions are the minimum stable feed, the usable range with the intended nozzle, metering consistency and whether the setting can be recorded.
Do not compare machines by maximum feed alone. Compare the range that produced an accepted result on the same contamination.
Trigger-on time and cleaning rate
Two crews can use the same feed setting and consume different quantities per shift because their trigger duty differs. Geometry, access, contamination thickness, operator technique and inspection pauses all change trigger time. Cleaning productivity also changes the denominator: consuming less per hour is not a saving if the job takes much longer or fails the acceptance test.
Record trigger-on minutes separately from total labor time. Both belong in cost analysis, but only trigger time belongs in the direct feed calculation.
Pellet condition, air and nozzle setup
Feed rate is only useful when the rest of the process is stable. Pellet size and condition, delivered compressed air, nozzle geometry, stand-off distance and angle affect removal. Degraded pellets, restricted air or an unsuitable nozzle can increase time and media use without improving the result.
If pellet delivery becomes intermittent, stop the consumption trial. A trial with unstable feed measures a fault, not the application. The pellet-feeding diagnostic guide defines that service boundary.
Delivery timing and sublimation
Dry ice loses mass while stored. Cold Jet states that the rate changes with climate and container insulation. That makes a universal loss allowance unsafe for purchasing. Ask the supplier for the delivered mass, container type, planned delivery time and handling requirements, then measure the loss across the actual interval between receipt and use.
Shorter storage can reduce loss, but delivery reliability also matters. The correct reserve covers a stated failure model—such as a delayed cleaning window or measured site loss—not an unexplained percentage added forever.
Measure kilograms per accepted output
A useful application trial records both process settings and material balance. Before starting, define the contamination, substrate, area or part count, cleanliness criterion, surface acceptance criterion and stop conditions. Then record:
- dry ice mass received and mass loaded;
- pellet size, delivery time and storage interval;
- machine model, feed setting and hopper refills;
- delivered air under load, nozzle, distance and angle;
- trigger-on time and total labor time;
- accepted output, rejected output and reason for rejection;
- dry ice remaining at the agreed end point.
Calculate consumed mass from the controlled starting and ending quantities. Divide it by accepted output. Repeat the trial on representative easy and difficult conditions rather than using only a freshly prepared demonstration surface.
The result should be a range with a documented setup, not a timeless constant. If contamination or geometry changes materially, that is a new application condition and should be revalidated.
Use consumption data to choose the system
Measured consumption supports better equipment questions. Check whether the accepted setting sits comfortably inside the machine’s stable feed range, whether the hopper permits practical refill intervals, and whether pellet size and nozzle options match the job. For automated work, confirm that the controller can reproduce feed, air, path and trigger timing rather than relying on an operator’s memory.
Cost should be expressed per accepted output. Include delivered pellets, measured storage loss, compressed air, labor, setup, inspection, rejected work and production interruption. Pellet price alone cannot establish operating cost, and machine price alone cannot establish payback. Local price and utilization are change axes; the accepted cleaning specification remains fixed.
Keep the safety boundary outside the calculation
Dry ice becomes carbon dioxide gas. NIOSH and OSHA publish occupational exposure and emergency-response boundaries for carbon dioxide. Ventilation, exposure assessment and any required monitoring must therefore be defined for the actual work area. A lower media rate does not remove the carbon-dioxide hazard, and a lower consumable cost does not justify exceeding a safe operating boundary.
Do not optimize storage by sealing dry ice into an unapproved container or by moving it into a poorly ventilated occupied space. Follow the dry ice supplier’s container and handling instructions and the site’s risk assessment. Stop the trial if ventilation controls are unavailable, exposure conditions are uncertain, feed is unstable or the surface no longer meets acceptance criteria.
For the first order, use the named machine’s feed range, realistic trigger time and measured site loss. For every later order, use dry ice per accepted output. DryIceSys can help define a machine configuration and trial record through the project enquiry form.