See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/282633853

Efficient use of compressed air for dry ice blasting

Article  in  Journal of Cleaner Production · July 2015

DOI: 10.1016/j.jclepro.2015.07.053

CITATIONS READS

7 157

2 authors, including:

Vítězslav Máša
Brno University of Technology
20 PUBLICATIONS   41 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Energy Savings for Small and Medium Enterprises View project

Strategic Partnership for Environmental Technologies and Energy Production View project

All content following this page was uploaded by Vítězslav Máša on 29 August 2019.

The user has requested enhancement of the downloaded file.

 Efficient Use of Compressed Air for Dry Ice Blasting
Vítězslav Mášaa, Pavel Kubaa

aBrnoUniversity of Technology, NETME Centre & Inst. of Process Engineering – VUT ÚPEI.
Technická 2896/2, 616 69 Brno, Czech Republic, masa@fme.vutbr.cz, tel. +420 541144916

This work was published at Journal of Cleaner Production

Please cite the work if it has been useful to you.
The article can be cited as:
Máša V., Kuba P., 2016, Efficient use of compressed air for dry ice blasting,
Journal of Cleaner Production, Vol.111, No.A, pp.76-84, ISSN 0959-6526, DOI:
10.1016/j.jclepro.2015.07.053
 Efficient Use of Compressed Air for Dry Ice Blasting
Vítězslav Mášaa, Pavel Kubaa

aBrnoUniversity of Technology, NETME Centre & Inst. of Process Engineering – VUT ÚPEI.
Technická 2896/2, 616 69 Brno, Czech Republic, masa@fme.vutbr.cz, tel. +420 541144916

Abstract

Dry ice blasting is a highly efficient and environmentally-friendly method for cleaning of contaminated
surfaces. The method is widely used in many industries, such as automotive, machinery and food
industry. Dry ice blasting is a technology that requires large amounts of compressed air which is
among the most expensive forms of energy currently employed in industries. Compressed air (air of
high pressure and volume) accelerates particles of dry ice during the blasting. Our paper presents a
comprehensive overview of saving measures which can greatly help decrease energy intensity of the
blasting.

Technology of dry ice blasting comprises four key elements: a compressor (generator of compressed
air), a blasting machine, transport of pellets, and a blasting nozzle. Our team successively analysed
these four key aspects of the technology and designed relevant savings measures, which results in
identification of minimum, theoretically attainable energy consumption. Experiments and operational
tests proved that application of all available savings measures may decrease energy consumption by
87 % compared to current industry standards.

It is possible to evaluate current technologies for dry ice blasting in terms of performance of the
blasting machine and its functional properties. So far, there has been no method for evaluation of
energy parameters of the technology. This paper presents a novel quantity which helps quantify and
evaluate energy consumption of dry ice blasting. Main benefit of the quantity called “specific energy
consumption of dry ice blasting” is the fact that it allows to compare energy consumptions of various
blasting systems. For the purposes of this paper, the novel quantity is especially used to quantify the
impact of recommended savings measures.

Keywords

dry ice blasting, compressed air, specific energy consumption, energy savings,

1. Introduction

Dry ice blasting is a cleaning process which accelerates particles of dry ice using compressed air; a
stream of these particles is directed at a contaminated surface and cleans it. Major advantages of dry
ice blasting include speed of the cleaning process and zero waste products (all the dry ice sublimes
once it hits surface). Although carbon dioxide (CO2) emissions are higher compared to other
technologies (Millman and Giancaspro, 2012), the process is not a great burden for the environment.
Carbon dioxide is standard by-product of chemical industry and its application in cleaning is basically a
recycling method (Spur et al., 1999). Membrane separation for the post-combustion CO2 capture was
reviewed by Khalilpour et al. (2014). No new carbon dioxide is released into the atmosphere during dry
ice production or use, so it does not increase CO2 concentrations in the atmosphere.
Under atmospheric conditions, carbon dioxide is a colourless, chemically inert gas. It does not induce
corrosion and does not adhere to solids. Dry ice is commonly transported liquefied in pressure vessels
at 12-20 bar (1.2-2 MPa). Dry ice refers to solid state of the matter which is reached at temperatures
lower than -78.5 °C. Snow-like dry ice is generated by expanding carbon dioxide through a nozzle.
Generated “snow” may be directly used in a blasting machine but dry ice pellets are much more
common. The pellets are little pins of pressed snow-like dry ice, with 3 mm diameter and 5-15 mm
length (Fig. 1a) which are generated in the so called pelletizer machine. If we further press the dry ice,
we obtain large-sized blocks which are more stable over time, can be easily transported, and may be
further processed (Fig. 1b).

Fig. 1. Pellets (a) and block (b) of dry ice

Dry ice blasting uses kinetic energy of accelerated dry ice particles hitting the surface. Dry ice
produces thermal shock to the undercooled material and causes embrittlement of the contaminated
layer. Dry ice further expands during sublimation of carbon dioxide (Uhlmann et al., 2009). Effects of
dry ice blasting may be attributed to the compressed air which accelerates both the particles of dry ice
and particles of the contaminated layer. Output rate of compressed air from an industrial blasting
machine exceeds 150 m/s. Quality features of dry ice include purity (carbon dioxide concentrations),
shape and age also have impact on efficiency of the cleaning.

The technology is widely used in all types of industries, such as automotive and foundry industries,
mostly for cleaning of functional surfaces and removing undesired materials (such as dust, rust, resins
and all sorts of grease). One of the specific features of dry ice blasting is a combination of the above
mentioned effects of dry ice and its low hardness which allows to efficiently remove the contaminants
without damaging the surface. Technology is very successful in cleaning various types of molds and
devices operating in high-temperatures. High temperature difference has a positive impact on the
cleaning process. In contrast to common cleaning with water, the equipment to be cleaned does not
have to be turned off and cooled. Certain applications in food industry use the same effect of the
technology to remove residues of dough from baking molds, forms and pans. Millar (2001) argues that
dry ice blasting further has a positive impact on elimination of certain types of bacteria. Promising
areas of utilization include cleaning of air units, treatment of materials before coating (painting and
varnishing), and special direct application on bearings surface. Besides blasting, dry ice cools foods in
food industry, and even substitutes ammonia in leather industry (Sathish et al., 2013).

Compared to other cleaning methods (water blasting, sand blasting, and blasting using a sand-like
material), dry ice blasting can be a less energy-intensive and time consuming technology (Millman and
Giancaspro, 2012). As to the efficiency of the cleaning, noise levels and operators’ skills, dry ice
blasting scores similar to the other cleaning methods above. Still, dependency of the technology on
compressed air supplies adversely affects its potential. Compressed air transports and accelerates dry
ice pellets, and helps corrupt the contaminants on the treated surface. However, compressed air is an
expensive medium due to a low efficiency of compressors (Dindorf, 2012). Systems working with
compressed air usually waste up to 80-93 % of supplied energy in the form of heat and/or losses
caused by leakages (SEI, 2007), Fig. 2. Volumetric flow rate of 6 bar (0.6 MPa) compressed air is
commonly about 5 Nm3 per minute (300 Nm 3/h). In order to achieve these parameters, it is necessary
to use powerful industrial compressors. Although consumption of compressed air decreases
worldwide, its generation still requires lots of energy. In Europe, generation of compressed air is
responsible for more than 10 % of electrical energy consumption in all industries (Radgen, 2005).

Fig. 2. Losses in systems using compressed air (SEI, 2007)

Efforts to decrease compressed air consumption in dry ice blasting processes are in line with current
worldwide trends of energy savings in industry. Tonn et al. (2014) thoroughly discusses various
approaches to energy savings. Jovanovic et al. (2014) examines a topic closely related to dry ice
blasting, that is production and distribution systems for compressed air with special focus on its
application and reliability in large-scale facilities. Although dry ice blasting technology has a great
potential for industrial cleaning, there are only few researchers who analyse and study this topic. Liu et
al. (2012) analyses cleaning mechanisms of dry ice blasting, Uhlmann et al. (2009) introduces new
fields of application and Foster (2012) discusses cleaning rate and costs of the dry-ice blasting.
However, our team has not found any papers dealing with the energy savings measures in the area of
dry ice blasting.

Essential benefit of our research is an introduction of a quantitative analysis of blasting systems’

energy intensity. So far, there has been no method to compare energy consumption of various dry ice
blasting machines and systems. Thanks to a newly introduced quantity called “specific energy
consumption of dry ice blasting” we are now able to evaluate the energy intensity and consumption.
Major contribution of our paper is an overview of energy saving measures for blasting system
elements (compressor, dry ice blasting machine, transport of dry ice and its acceleration). So far, no
one has published this type of an overview. Minimum attainable energy consumption of dry ice
blasting is also discussed in the paper.

2. Dry Ice Blasting Machine

A dry ice blasting machine is a relatively simple piece of equipment that prepares dry ice particles for a
transport to the contaminated surface. There are only few companies offering this technology
worldwide. And moreover, design of these machines is very similar. All the machines use compressed
air for acceleration of the dry ice. Industrial applications work with two basic concepts. First, there is
the so called “block shaving system” (Fig. 3) using dry ice blocks (250 x 125 x 125 mm). Compared to
dry ice pellets, dry ice blocks have longer storage life. The blocks are shaped in the machine using a
shaving mechanism. Particles of dry ice are shaved from the dry ice blocks using a rotating knife
wheel. Particles size depends on the knife wheel rotations, i.e. on rotations of a pneumatic motor
which drives the wheel. Shaved dry ice is lead to an airlock device which dispenses an appropriate
amount of dry ice through the pressurized hose with a Laval nozzle mounted at the end.

Fig. 3. Dry ice blasting machine with a block shaving system

Second, there is the so called “pellet system” producing dry ice pellets (Fig. 4). The pellets are
supplied into the airlock device without any previous shaving. The airlock device feeds dry ice pellets
into a stream of compressed air. This technological concept does not allow for any variations in the
size of the dry ice particles. The pellet system is a simpler, yet less versatile dry ice blasting machine
than the block shaving system.

Equipment using liquid carbon dioxide directly from a pressure vessel is a special alternative to the
blasting technology. Snow-like dry ice is generated by expansion of carbon dioxide through the nozzle,
compressed air accelerates the dry ice. The technology helps clean fine surfaces, such as electronic
components and circuits.
Fig. 4. Dry ice blasting machine with a pellet system

Output rate of dry ice particles has a decisive impact on efficiency of the blasting process. Output rate
starts at 80 m/s, depending on a type of application, and may theoretically exceed the speed of sound
(Ma > 1). Acceleration of large amounts of pellets to such a speed obviously requires compressed air
with high pressure and flow rate. In order to thoroughly analyse the blasting system in terms of energy
parameters, following information and data must be available:

 Range of compressed air operational pressure (in catalogues as Blasting pressure), in bar
 Consumption of compressed air, in Nm 3/h
 Consumption of dry ice (or feed rate of CO2), in kg/h

Flow rate of compressed air is related to the amount of transported dry ice. Our team researched dry
ice blasting machines available on the market worldwide now and drafted a table of a characteristic
range of these quantities (Tab. 1). We especially used the parameters which are available on the
websites of two world´s leading producers from the United States (Cold Jet, 2015) and Switzerland
(ASCO, 2015).

Quantity Common range

Range of compressed air operational pressure (in technical
specifications as Blasting pressure) 1 – 20 bar
Consumption of compressed air (in technical specifications as Air
consumption) 18 – 420 Nm³/h
Consumption of dry ice (in technical specifications as Feed rate of
CO2) 20 – 190 kg/h

Tab. 1 An overview of a common range of operational quantities

A size of a commercial blasting machine does not exceed 1,100 x 650 x 1,200 mm (LxWxH) and
weight does not exceed 180 kg. These maximum values are common for the machines with the
highest performance. These also have the highest potential for energy savings thanks to their wide
use in commercial applications.

3. Novel Quantity for Evaluation of Energy Intensity of Dry Ice Blasting

Energy consumption of the dry ice blasting system is associated with its performance. In order to
compare energy consumptions of various systems, our team decided to introduce a novel quantity.
The quantity describes efficiency of the utilization of compressed air for acceleration of dry ice to a
required speed. It relates consumed energy in kWh to an amount of blasted dry ice in kg. Operators of
blasting machines usually have information about consumption of dry ice, and are therefore able to
compare efficiency of particular machines using this quantity. The quantity is called a specific energy
consumption of dry ice blasting EB, and is measured in kWh/kg.

𝑄𝑎 ∙ 𝑆𝐸𝐶
Eq. 1. 𝐸𝐵 = 𝑄𝑖

where Qa is a consumption of compressed air in Nm 3/h, Qi is a consumption of dry ice particles in kg/h,
and SEC is specific energy consumption for generation of compressed air in kWh/Nm 3 (Dindorf, 2012).

𝑃𝐴
Eq. 2. 𝑆𝐸𝐶 = 𝑞𝐹𝐴𝐷

where PA is average power input of compressor (kW) and qFAD is free air delivery (Nm 3/h).

For most cleaning applications using blocks of dry ice (Fig. 3), common specific energy consumption
of dry ice blasting EB reaches ca. 0.36 kWh/kg. We calculated this value based on Eq. 1 for system
parameters given in Tab. 2. In order to be able to compare blasting machines, we must fix the air
pressure (as it affects SEC) and output rate of a mixture of compressed air and dry ice. The
parameters given in table 2 were determined upon consulting with businesses operating in the area of
dry ice blasting, and they are deemed to be the most common values attained in the process of
cleaning with dry ice. We shall consider these values for our further analysis to be standard.

Quantity Value Unit

Compressed air operational pressure 6 bar
Consumption of dry ice (Qi) 100 kg/h
Consumption of compressed air (Qa) 300 Nm³/h
SEC of compressor 0.12 kWh/Nm3
Particles output rate 160 m/s

Tab. 2 Standard operational parameters of a dry ice blasting machine for industrial cleaning

EB of a pellet system (Fig. 4) commonly reaches ca. 0.31 kWh/kg, since the consumption of
compressed air Qa of a pellet system is lower (the pellet system has no shaving mechanism which
requires ca. 40 Nm3/h , see chapter 4.2). EB = 0.31 kWh/kg is therefore based on Eq. 1 after Qa = 260
Nm3/h is established. Operational pressure for both the shaving mechanism and pellet system is 6 bar
(0.6 MPa).

4. Minimisation of Compressed Air Consumption

Minimization of compressed air consumption is a basic measure for decreasing energy intensity of dry
ice blasting. In order to minimize the consumption, we identified all points where compressed air is
employed. Then we drafted an overview of measures available for decrease in compressed air
consumption in dry ice blasting. We discuss particular parts of the technology (Fig. 3) and relevant
potentials for savings. We experimentally verified minimum attainable compressed air consumption
and assumed that the compressed air would be used only for transport of dry ice and not for its
acceleration (or preparation in the block shaving system). Therefore we measured minimum flow rate
and pressure of compressed air necessary for transport of dry ice pellets in the hose. We present a
detailed description of the experiment in chapter 4.3 and relevant consequences for minimum
consumption in chapter 4.5.
 4.1. Supply of Compressed Air

We wish to start with a supply of compressed air. Compressed air is generated by various types of
compressor units. The units must have sufficient parameters (Tab.1). Piston compressors are the
most commonly used machines since they require the lowest investments, and are able to supply air
for all operational modes of the blasting machine. Industrial applications often employ screw
compressors as well as rotary vane compressors, too. Their noise level is lower, and the design is
simpler (more robust) compared to the piston compressors. However, repairs on the machines are
rather complicated. High performance tasks usually require compressors with combustion engines.

Energy consumption (mostly consumption of electrical energy) is affected by a type of the compressor
as well as its age (Dindorf, 2012). SEC, specific energy consumption mentioned above (Eq. 2), is a
universal tool for comparison of individual compressors. Unfortunately, there is no summary study
comparing SEC for various types of compressors.

Dindorf mentions SEC and its significance, and presents values for piston compressors with 7 bar
operational pressure. Very good SEC values in modern compressors should range from 0.085 to 0.11
kWh/Nm3. SEC in most of the common compressors ranges from 0.11 to 0.13 kWh/Nm 3. Machines
with a higher SEC should be replaced or upgraded. Difference between a standard and a modern
supply of compressed air may reach up to 33 %.

One of the main options to decrease consumption of compressed air is to control the compressor and
adjust its performance to current requirements. Compressor output pressure may be decreased to the
pressure level necessary for blasting. Compressor output pressure must be controlled. Two most
common features include two-state operation control and continuous speed control. Compressors with
two-state operation control turn off after reaching a sufficient pressure, and turn on if the pressure
drops below a specific level. However, only small compressors which are not frequently turned off and
on again may be controlled this way. Continuous speed control is a more efficient control method with
lower energy consumption. Desired parameters of the pressure, especially in electric drive
compressors, may be set using a frequency converter, which prevents unnecessary idling, i.e. over-
pressurization of the air tank. If we use this kind of a system, we may decrease consumption of the
electrical energy by 60 % (Dizadji and Rashtchi, 2014).

4.2. Dry Ice Blasting Machine

Most of the compressed air in a dry ice blasting machine serves as a medium for acceleration of dry
ice particles. However, in case of the block shaving system, significant share of the compressed air is
consumed in a pneumatic motor of the shaving mechanism. Repeated experimental measurements of
a well-maintained, modern machine provide values of 37.40 ± 0.20 Nm3/h. Consumption may be
higher in older, poorly-maintained machines (Máša et al., 2014). Consumption of 37 Nm 3/h amounts
for ca. 12 % of a total consumption of the dry ice blasting system (300 Nm3/h).

Our research team has introduced an innovation which basically eliminates this energy consumption.
Replacement of a pneumatic drive of the shaving mechanism with an electric drive is an ideal saving
measure. Basic parameters for replacement of the existing pneumatic drive are its performance and
rotations. Asynchronous, three-phase 0.55 kW, 50 Hz frequency electric motor is a good alternative to
the existing drive. The electric motor proved it was a suitable drive for the shaving mechanism.
Excellent properties of the motor were repeatedly tested in a well-maintained machine under
operational conditions.
Compared to a pneumatic drive, electric drive has significantly lower levels of energy consumption.
Consumption of pneumatic drive of 37.4 Nm 3/h corresponds with power input of the compressed air
supply of 4.5 kW (piston compressor with SEC = 0.12 kWh/Nm 3). 4.5 kW thus corresponds to 100 %
block shaving system energy consumption. New electric drive has a rated output of 0.55 kW.
Experiments proved that actual consumption of the drive during operations (dry ice shaving) equals
only ca. 0.17 kW. We could thus install a drive with a lower performance. Replacement of a pneumatic
drive with an electric drive may save up to 4.3 kW, i.e. 96 % of energy.

In order to control motor rotations (and thus the dry ice particles size, too), frequency converter must
be integrated into the motor. In terms of serviceability, properties of the new electric motor and the
original pneumatic motor do not significantly differ. Even the machine operator does not face any
difficulties. It is only necessary to ensure stable power supply during the blasting. It should not be
difficult with the one-phase power supply of the motor with frequency convertor. This innovation is a
great contribution in all facilities where the insufficient flow rate of the compressed air may prevent
application of the dry ice blasting technology (or where a more powerful compressor would have to be
leased instead).

Supply of compressed air in both of the analysed systems is further necessary for an airlock device
(Fig. 3 and 4). Thanks to a rotary motion of an impeller with slots, dry ice is evenly transported from a
hopper into a hose. There have been no experiments testing the consumption of compressed air in an
airlock device drive and we therefore included it in the total consumption of compressed air. Electric
motor with a frequency converter is an economically more beneficial option for the airlock device drive.
Principles of energy savings are similar to those of the shaving mechanism. Electric drive is a common
part installed in blasting machines. Power input of 0.5 kW is enough for pellet flow rate of 100 kg/h.

4.3. Transport of Dry Ice

Dry ice (pellets or particles) is transported only through a hose in a stream of compressed air (Fig. 5a).
Single-hose transport system is currently the most used machine arrangement. Compressed air
streams from the compressor and dry ice is fed into the stream via the airlock device. The mixture of
air and dry ice is delivered to the nozzle (Fig. 3 and 4). Pressure in the transport hose is equal to the
maximum pressure of the blasting machine and this may therefore cause cracking of the pellets. This
disadvantage is, however, compensated by a simple configuration of the machine and easy
manipulation for the operator.

Two-hose transport system with a jet-pump is an older transport configuration. Compressed air
streams in one hose (the so called acceleration hose), and dry ice streams in the other one (transport
hose). Both hoses are connected in a blasting gun with a nozzle. Air flow from the acceleration hose
creates negative pressure in the transport hose. Speed of the dry ice pellets in the transport hose is
lower, which ensures that the pellets are not damaged before reaching the nozzle (Stratford, 2000).

However, since the more simple one-hose systems are more common in cleaning applications, we will
further discuss this type of blasting machine arrangement. Consumption of energy necessary for
transport of dry ice in the hose is an interesting issue. If we isolate the hose from the rest of the
blasting system, we can identify minimum pressure and flow rate of compressed air for transport of dry
ice. We measured a 20 m hose with an inner diameter of ¾” (1.905 cm), which are the most widely
used dimensions in cleaning applications. Compressed air was supplied from a compressor with a
volumetric flow rate of 60 Nm 3/h and maximum inner pressure of 9 bar. In order to decrease the
pressure, we worked with 0-16 bar pressure regulator with a “compressor–filter–regulator”
configuration. Pellets of dry ice are fed into the system using a 100 kg/h capacity hopper.
Experiments identified consumption and pressure of the compressed air, and mass flow of pellets in
the feeder and hose. Consumption of compressed air was measured via an orifice plate (Fig. 5b.1)
with a datalogger (Fig. 5b.2). Absolute pressure (Fig. 5b.3), differentiation pressure (Fig. 5b.4) and
temperature (Fig. 5b.5) of streaming medium were measured and used in calculations of volumetric
flow rate of the compressed air. Pressure regulator allows to control accurate setting of operational
pressure in the hose. By lowering the pressure, we could reach the minimum pressure and flow rate
while maintaining successful transport of dry ice pellets to the hose output. We observed that
compressed air flow rate of 18.5 Nm3/h and absolute pressure of 1.13 bar (0.113 MPa) was necessary
for transport of 100 kg of pellets in an hour over 20 m distance, which corresponds with electrical
power input of compressed air supply of 1.57 kW (piston compressor with SEC = 0.085 kWh/Nm3).

Fig. 5. Dry ice pellet stream on the end of the hose (a) and measurement of compressed air
consumption by orifice plate (b)

Screw conveyer may theoretically replace the compressed air for transport, which is yet another option
for energy savings in dry ice blasting. Conveyer should be equipped with a dry ice hopper, drive and
transport pipe itself. Screw conveyer with these parameters must be powered by a source with power
input P = 0.25 kW (Eq. 3.) (Kulinowski and Kasza, 2015)

𝜆 ∙ 𝐿+𝐻 𝐷∙𝐿
Eq. 3. 𝑃 = 𝑄𝑡𝑖 ∙ +
367 20

where Qti = 0.1 t/h is a mass flow rate of transported dry ice pellets (equivalent of consumption of dry
ice in Eq. 1), λ = 2.5 is a coefficient of pellets resistance, L = 20 m is a conveyor length, H = 0 m is a
change to a conveyor height, D = 0.15 m is a conveyor diameter.

Compared to the hose system, electrical energy savings reach approximately 80 %. However, we will
dispense this option since mechanical conveyers are impractical. Cleaning with dry ice must be
flexible and fast since the machine often changes its position. Mechanical conveyers are not flexible
and their mobility is much worse compared to the hoses. Therefore they fail to comply with basic
requirements on practical transport of dry ice.

If we reduce transport distances of dry ice and thus also decrease compressor performance (pressure,
air flow rate), we are able to save significant amounts of energy.
 4.4. Acceleration of Dry Ice

Kinetic energy of particles hitting the contaminated surface is a core element in efficiency of dry ice
blasting. Acceleration is responsible for the kinetic energy. Kinetic energy Ek (in J) of the particles is
defined by an equation:

1
Eq. 4. 𝐸𝑘 = 𝑚 ∙ 𝑣2
2

where m is weight of dry ice particle in kg, and v is velocity of dry ice particle in m/s. If we assume that
the particle size (and therefore their weight, too) is optimum for a given type of cleaning, the only
parameter that may be adjusted is the particle velocity.

Nozzles

The pellets or granules are usually accelerated by the compressed air stream coming through a nozzle
at a jet pressure of 3 to 12 bar (0.3-1.2 MPa) and a maximum volume flow rate of 11 Nm 3/min,
depending on the application in question and depart from the nozzle at near supersonic speed (Spur
et al., 1999). A Laval nozzle provides the highest acceleration of air flow for a given type of streaming
(Fig. 6). Laval nozzle has convergent and divergent sections. Air velocity at the narrowest part of the
convergent section (throat) may become sonic (Ma = 1). Ma = 1 requires a particular ratio of pressure
prior to and beyond the nozzle (pc/pb). Increasing the ratio results in increase in mass flow in the
nozzle. But further increase in the ratio cannot increase mass flow anymore. However, velocity in the
divergent section may be increased above Ma > 1 (depending on pc/pb ratio).

Fig. 6. Laval nozzle (engApplets, 2014)

Laval nozzles are currently the most used type of nozzle for dry ice blasting. But for optimum
acceleration and even distribution of particles, non-circular shape of nozzles is better. Trajectory of
streaming particles should not be changed, as proved by Lehnig’s patents (1997). If we do change the
trajectory, particles may concentrate in the middle of the stream, and area of efficient blasting will
reduce. Lehnig (1997) presents a non-symmetrical nozzle design which changes reflection of potential
shock waves generated at the nozzle throat generating Mach lines (or expansion lines). Benson
(2008) provides basic information about generation and properties of these specific shock waves.

Nozzles must above all efficiently utilize the compressed air. Bad nozzle prolongs the cleaning
process and thus raises operational costs. Another thing to consider in nozzles is their length since the
longer the nozzle, the more the dry ice particles become damaged as the particles more frequently hit
the wall and each other along the way out. On the other hand, a long nozzle provides sufficient time
for particle acceleration.

Mechanical acceleration

A rotor jetting the dry ice pellets is an alternative to the nozzle. Pellets are accelerated and axially
propelled to the centre of a rotating disc. Dry ice pellets receive sufficient kinetic energy thanks to a
centrifugal force (and not from a high volumetric flow rate of compressed air). Rotor has a high
peripheral speed from a suitable electric motor. The sole source of energy for pellets acceleration is
electrical energy transformed into mechanical energy and transmitted via a shaft to the rotor. We may
call this system a centrifugal accelerator.

It refers to a patented solution “Centrifugal accelerator, system and method for removing unwanted
layers from a surface“ (Foster and Fisher, 1995). The accelerator was originally designed for transport
of cryogenic materials (deuterium/tritium) into fusion reactors. Identical machine was designed for
elimination of coating from airplanes. Experiments with dry ice pellets proved that the rotor has a
similar cleaning effect as the conventional blasting machines using compressed air, but the rotor has
lower energy consumption (Foster et al., 1995). Accelerator has its own device for generation of round
dry ice particles, Fig. 7a. The system is further equipped with infrared lamps which supply heat for
preheating of the contaminated surface. Experiments show that preheating of coated surfaces greatly
enhances efficiency of the process, Fig. 7b (Foster et al., 1995).

Foster

Fig. 7. Centrifugal accelerator with pelletizer unit (a) and surface heater (b)

tested a machine with a disc rotor of 350 mm in diameter. 11 kW brushless DC electric motor drove
the rotor and maintained 12,000 rotations per minute (rpm). A mass flow of pellets entering an inlet
hole reached approximately 454 kg/h. Speed of the dry ice pellets leaving the centrifugal accelerator
was measured using a high-speed camera. Average output rate of pellets was 350 m/s.

Another machine for acceleration of pellets using centrifugal force was patented in 2008. Institute for
Applied Research in Fraunhofer (Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung)
along with Technical University Berlin is the original assignee. The patent no DE102006029437 is
called “Vorrichtung und deren Verwendung zum Schleuderstrahlen mit Trockeneis“ (Mernissi et al.,
2008), Fig. 8.
Machine has a similar principle as abrasive blasting machines which utilize sand or metal balls.
Machine is equipped with an impeller with curved blades, where dry ice pellets axially enter the
impeller. The pellets are transported from the hopper, which is close to the impeller, via a screw
conveyor.

The impeller with the curved blades has a 400 mm diameter and is driven by an electric motor which
maintains 2,400 rotations per minute. A high-speed camera was used to measure the output rate of
dry ice pellets from centrifugal accelerator and obtained the value of 62.8 m/s. This is a very low speed
and insufficient for most of cleaning applications. In order to reach supersonic speed, impeller must
have much higher rotations than it has now.

Fig. 8. Model of centrifugal accelerator (Uhlmann et al., 2009)

Unfortunately, accelerators have not become very popular in practice. The main reasons are their
larger dimensions compared to nozzles, and requirements for high rotations of the rotor (over 10,000
rpm) so that dry ice is sufficiently propelled. Yet in terms of energy intensity, accelerators are a
promising concept which definitely deserves further research. Our calculations prove that 2 kW motor
is enough to provide kinetic energy for propelling the dry ice. Calculations worked with a rotor of 250
mm diameter and parameters of a typical blasting machine (Tab. 2).

4.5. Results and Discussion

Results of the energy savings assessment of the blasting machine individual parts help identify
minimum attainable energy consumption of the blasting machine as a whole. We may recommend
following procedures:

In block shaving systems, replace a pneumatic motor of the shaving mechanism with an electric
motor. This saves up to 12 % of air consumption. If the blasting machine is equipped with a modern,
efficient compressor whose SEC is 0.085 kWh/Nm3, EB may reach approximately 0.22 kWh/kg which
is a 38 % saving compared to conventional systems with EB equal to 0.36 kWh/kg (Fig. 9).
In pellet systems, we can obtain best energy savings if we replace the conventional nozzle with a
centrifugal accelerator. This decreases attainable EB to 0.041 kWh/kg using a combination of various
saving measures. We are not considering here a shaving mechanism since the system uses dry ice
pellets. Compressed air carries dry ice only from the blasting machine to the accelerator. Final values
of EB (kWh/kg) may be identified using an equation 5 which is an application of equation 1:

𝑃𝐶 + 𝑃𝐶𝐴 + 𝑃𝐴𝐷
Eq. 5. 𝐸𝐵 = 𝑄𝑖

where PC is a compressor power input, PCA is a power input of a centrifugal accelerator drive, PAD is a
power input of a pellet feeder electric motor, and Qi is a consumption of dry ice pellets in kg/h. A small,
mobile compressor with PC = 1.6 kW is able to suffice required compressed air, which has been
verified by experiment presented in chapter 4.3. Calculated PCA is 2.0 kW (chapter 4.4). Expected
consumption of a feeder drive is PAD = 0.50 kW (chapter 4.2). Configuration of this best energy saving
system is displayed in Fig. 10.

0,40

0,35

0,30 38 %
Specific Energy Consumption [kWh/kg]

87 %
0,25

0,20

0,15

0,10

0,05

0,00
Conventional With Electric Motor and Conventional With Accelerator and
Efficient Compressor Unit Efficient Compressor Unit
Block Shaving System Pellet System

Fig. 9. Specific energy consumption of both blasting systems and attainable savings

Particular measures that are presented in the paper may significantly decrease consumption of
compressed air. EB = 0.041 kWh/kg is a minimum attainable value of specific energy consumption of
dry ice blasting. In order to achieve this value, use of dry ice pellet accelerator instead of a nozzle is
necessary. However, accelerators are not available on the market anywhere in the world. Obstacles to
their expansion is a more complicated design compared to simpler and smaller nozzles. Yet, the
accelerators are crucial in energy savings. A pellet system with an accelerator (0.041 kWh/kg) is 87 %
more efficient than a conventional pellet system (0.31 kWh/kg), as obvious in Fig. 9. Broad application
of accelerators in industries is conditioned by their future development.
Fig. 10. Installation diagram of energy efficient blasting unit

5. Conclusion

Operators of blasting machines agree that costs of compressed air production greatly limit application
potential of dry ice blasting technology. Efforts aimed at decreasing consumption of the compressed
air cannot focus on production costs only. Many commercial facilities usually do not have a powerful
compressor and an operator of a blasting machine has to lease a high performance mobile
compressor and have it transported to the premises. This greatly increases costs of the service.
Particular measures that we presented here may significantly decrease consumption of compressed
air.

Our paper discussed savings measures which may be implemented by adjusting the design and
technology of blasting machines. Experiments and operational tests proved that application of all
available savings measures may decrease energy consumption by 87 % compared to current industry
standards. We have not dealt with the issue of cleaning efficiency but there is just as much potential
for energy savings and losses. If the machine blasts 100 kg of ice in an hour, it does not necessarily
mean that the cleaning is efficient. It is the operators who are responsible for optimum use of dry ice.
First, they have to plan the cleaning process carefully so that the same surface area is not repeatedly
cleaned and the cleaned area is not soiled again. Second, they must find an optimum distance
between the nozzle and the surface, and a proper speed for moving along. Even an angle of a nozzle
tilt is important. Parameters of the blasting machine must be set to correspond with the degree of
contamination (especially operational pressure of compressed air, pellet output rate and flow rate).

We introduced a new specific quantity for comparison of various blasting systems; the quantity relates
consumed energy in kWh to an amount of blasted dry ice in kg. Operators of blasting machines
usually have information about consumption of dry ice, and are therefore able to compare efficiency of
particular machines thanks to specific energy consumption. Specific energy consumption of dry ice
blasting may be very helpful for manufacturers of blasting machines in their future innovations.

In real-life situations, we always have to consider whether an optimization of cleaning efficiency is

justifiable compared to the results, or not. Commercial facilities often clean contaminated surfaces
during shutdowns. Speed of the cleaning might therefore become more important than energy savings
and/or dry ice savings.

In this paper we presented general possibilities of decreasing energy intensity of dry ice blasting
machines. Operators will permanently benefit from the adjustments and innovations. Every time the
blasting machine is used, they save energy and operational costs.
6. Future Work

Acceleration of the dry ice pellets has the highest potential for energy saving. We wish to relate our
work to the published patents and develop an accelerator with a novel design. The new accelerator
will meet requirements on a wide spectrum of applications and will use less compressed air. The most
important task is to minimize dimensions of the accelerator and prolong its life. Successful operations
of this accelerator would bring enormous savings of compressed air and thus large savings of energy
and operational costs.

Second, we wish to research a process of dry ice degradation. Quality of dry ice greatly affects
efficiency of the cleaning process. Experiments proved that dry ice gradually sublimes which produces
micro-cracks, and the dry ice pellets may further freeze together under wrong storage conditions. This
obviously has a negative impact on their mechanical properties. Aging process (degradation) of dry ice
has not been researched and sufficiently analysed yet.

Acknowledgement

This work is an output of research and scientific activities of NETME Centre, regional R&D centre built
with the financial support from the Operational Programme Research and Development for
Innovations within the project NETME Centre (New Technologies for Mechanical Engineering), Reg.
No. CZ.1.05/2.1.00/01.0002 and, in the follow-up sustainability stage, supported through NETME
CENTRE PLUS (LO1202) by financial means from the Ministry of Education, Youth and Sports under
the „National Sustainability Programme I“.

Also, we would like to thank Mr. Dalimil Petrilák (Alkion service s.r.o., www.alkion.eu/en) for expert
consultation during experimental measurements and for providing us with blasting equipment and dry
ice.

List of references

 ASCO, 2015. Dry Ice Blasting Machines.

<www.ascojet.com/fileadmin/user_upload/ascojet.com/Dokumente/PDF/Produkte/Strahlgeraet
e/ASCOJET_Dry_Ice_Blasting_Machines.pdf> accessed: 13.4.2015
 Benson, T., 2008. Centered Expansion Fan. <www.grc.nasa.gov/WWW/K-
12/airplane/expans.html> accessed: 15.12.2014
 Cold Jet, 2015. Dry Ice Blasting Machines. <www.coldjet.com/en/products/dry-ice-blasting>
accessed: 13.4.2015

 Dindorf, R., 2012. Estimating Potential Energy Savings in Compressed Air Systems. Procedia
Eng. 39, 204-211 (www.sciencedirect.com/science/article/pii/S1877705812024125)

 Dizadji, N., Rashtchi, M., 2014. An Experimental Investigation of Energy Saving in Air
Compressors. Chem. Eng. Trans. 39, 151-156 (www.scopus.com/inward/record.url?eid=2-
s2.0-84908116149&partnerID=40&md5=fc3c5a183cdf8af67509c7350cc9ab8c)
 engApplets, Nozzle Applet. <www.engapplets.vt.edu/fluids/CDnozzle/cdinfo.html> accessed:
15.12.2014.
 Foster, C.A. and Fisher, P.W., 1995. Centrifugal Accelerator, System and Method for
Removing Unwanted Layers from a Surface. US Patent 5472369, Oak Ridge, United States.
 Foster, C.A., Fisher, P.W., Nelson, W.D., Schechter, D.E., 1995. A Centrifuge CO2 Pellet
Cleaning System. NASA Aerospace Environmental Technology Conference, 379-389
(hdl.handle.net/2060/19950025363)
 Foster, R.W., 2012. Carbon Dioxide (Dry-Ice) Blasting. <ecojetinc.com/ecopress/wp-
content/uploads/2012/10/RobertWFoster.pdf> accessed: 15.12.2014

 Jovanovic, V., Stevanov, B., Šešlija, D., Dudić, S., Tešić, Z., 2014. Energy Efficiency
Optimization of Air Supply System in a Water Bottle Manufacturing System. J. Clean. Prod.
85, 306-317 (www.sciencedirect.com/science/article/pii/S095965261400242X)

 Khalilpour, R., Mumford, K., Zhai, H., Abbas, A., Stevens, G., Rubin, E.S., 2014. Membrane-
based Carbon Capture from Flue Gas: a Review. J. Clean. Prod. 103, 286-300
(www.sciencedirect.com/science/article/pii/S0959652614010920)
 Kulinowski, P., Kasza, P., 2015. Screw Conveyors.
<www.kmg.agh.edu.pl/sites/default/files/pdf_ScrewConveyors_eng.pdf> accessed: 13.4.2015
 Lehnig, T.R., 1997. Nozzle for Cryogenic Particle Blast System. US Patent 5660580,
Cincinnati, United States.

 Liu, Y.H., Hirama, D., Matsusaka, S., 2012. Particle Removal Process during Application of
Impinging Dry Ice Jet. Powder Technol. 217, 607-613
(www.sciencedirect.com/science/article/pii/S003259101100653X)

 Máša, V., Kuba, P., Petrilák, D., Lokaj, J., 2014. Decrease in Consumption of Compressed Air
in Dry Ice Blasting Machine. Chem. Eng. Trans. 39, 805-810
(www.scopus.com/inward/record.url?eid=2-s2.0-
84908106018&partnerID=40&md5=844786cc2722530286b97d87e7d3d981)
 Mernissi, A.E., Gottheil, I., Krieg, M., Uhlmann, E., 2008. Vorrichtung zum Schleuderstrahlen
von empfindlichen Strahlmitteln, insbesondere Trockeneis (Device for Centrifugal Blasting of
Sensitive Abrasives, especially Dry Ice). DE Patent 102006029437, Berlin, Germany.
 Millar, I., 2001. Cold Jet - A Novel Technique for Cleaning & Decontaminating Food
Processing Areas, Equipment, Carcasses and Foods.
<www.foodbase.org.uk//admintools/reportdocuments/179-1-
313_B02006_ColdJet_REVISED_Final_Technical_Report_1t.pdf> accessed: 15.12.2014
 Millman, L.R., Giancaspro, J.W., 2012. Environmental Evaluation of Abrasive Blasting with
Sand, Water, and Dry Ice. Int. J. Archit. Eng. Constr. 3, 174-182
(www.iasdm.org/journals/index.php/ijaec/article/view/85/49)

 Radgen, P., 2005. Internet-Based Benchmarking for Compressed Air Systems.

<www.aceee.org/files/proceedings/2005/data/papers/SS05_Panel03_Paper12.pdf> accessed:
15.12.2014

 Saidur, R., Rahim, N.A., Hasanuzzaman, M., 2010. A Review on Compressed-air Energy Use
and Energy Savings. Renew. Sustain. Energy Rev. 14, 1135-1153
(www.sciencedirect.com/science/article/pii/S1364032109002755)

 Sathish, M., Madhan, B., Saravanan, P., Rao, J.R., Nair, B.U., 2013. Dry Ice – an Eco-friendly
 Alternative for Ammonium Reduction in Leather Manufacturing. J. Clean. Prod. 54, 289-295
(www.sciencedirect.com/science/article/pii/S0959652613002941)
 SEI, 2007. Compressed Air Technical Guide.
<www.seai.ie/Your_Business/Energy_Agreements/Special_Working_Groups/compressed_air
_SWG2007/Compressed_Air_Technical_Guide.pdf> accessed: 15.12.2014

 Spur, G., Uhlmann, E., Elbing, F., 1999. Dry-ice Blasting for Cleaning: Process, Optimization
and Application. Wear 233–235, 402-411
(www.sciencedirect.com/science/article/pii/S0043164899002045)
 Stratford, S., 2000. Dry Ice Blasting for Paint Stripping and Surface Preparation. Met. Finish.
98, 493-499

 Tonn, B., Frymier, P.D., Stiefel, D., Skinner, L.S., Suraweera, N., Tuck, R., 2014. Toward an
Infinitely Reusable, Recyclable, and Renewable Industrial Ecosystem. J. Clean. Prod. 66, 392-
406 (www.sciencedirect.com/science/article/pii/S0959652613007701)

 Uhlmann, E., Hollan, R., Mernissi, A., 2009. Dry Ice Blasting – Energy-Efficiency and New
Fields of Application, in: Pantelakis, Spiros, Rodopoulos, Chris (Eds.), Engineering Against
Fracture. Houten, Springer Netherlands, 399-409

View publication stats