In this study, first, the problems existing in jet pulsed filter (JPF) systems were listed, then the benefits of the new generation artificial intelligence-based jet filter process controller developed by Jetmaster company to eliminate all these problems, increase product quality and reduce energy consumption were explained.

Dr. Nihat Çankaya
Necmettin Erbakan University
Meram Vocational
School Department of
Food Processing
1. Introduction
Jet pulsed filter (JPF) is the most basic equipment used in flour, semolina, pasta, seeds, feed, bulgur, grain, and pulses industries. In all facilities where powder and granular raw materials are used or produced, dust collection and dust separation are done with JPF. Figure 1 shows the principle schematic of a JPF system [1]. JPF systems are included in all processes such as transportation, cleaning, grinding, sieving, sorting, and bagging in the mills. The problems experienced in these systems directly affect the production process. In current JPF applications, there is no solution that can easily detect their failures. However, many existing problems have become easily identifiable thanks to newly developed high-tech equipment and artificial intelligence-based intelligent control algorithms. In this study, first, the problems existing in JPF systems were listed, then the benefits of the new generation artificial intelligence-based jet filter process controller developed by Jetmaster company to eliminate all these problems, increase product quality and reduce energy consumption were explained.
2. Effects of Jet Filters on Product Quality
Although JPFs are among the most important pieces of equipment for the milling industry, they have remained in the background compared to other machines in today's milling. However, these filters must be working smoothly to catch the targeted values of main milling parameters such as grinding tonnage, ash ratio, and extraction rate. When the mill is first commissioned, while the jet filters are operating at full capacity, over time, the malfunctions in the filters appear as fluctuations in the ash value of which the reason cannot be understood. Then, the desired extraction rate in normal production tonnage becomes unachievable. This results in a reduction in production tonnage or extraction rate. In the same process, unit energy consumption also increases.

3. Current Problems
The jet filter system, which was developed by Bühler in 1979, is still widely used, and there are rarely personnel who have full knowledge of the working logic of this system. Disassembly and assembly of jet filter equipment takes a long time and makes fault detection difficult. In the studies carried out in flour and semolina mills, it has been observed that only 15% of the jet filters operate at their normal performance value, and the remaining 85% are used far below their performance. It has been determined that 20% of these filters do not work at all. Despite such a high-performance loss, it is very difficult to detect the malfunction that occurs because the filter areas are above the need. Problems in filters are always reflected in production. It is an essential requirement that all malfunctions be detected electronically, not by human eyes.
3.1. Problems Experienced in the Production Process Due to Jet Filter
In flour production, suction provided by the pneumatic fan must be at a constant value. In this way, the air velocity and the amount of load going to the passages could be realized at the planned values, so the targeted extraction rate could be obtained at the highest tonnage. Normally, a suction of around 100 mbar is provided by the pneumatics. When clean filter bags are first installed, a pressure drop of around 2 mbar occurs and a suction of around 98 mbar is transferred to production. However, over time, the filter bags begin to fill up and the pressure drop reaches 25 mbar levels. In this case, the suction that could be transferred to production was decreased at 75% and it drops 75 mbar. This decrease causes a decrease in tonnage. If the same tonnage is insisted on, then the ash value will increase and the extraction rate will decrease. As an alternative solution, it is seen that the fan speed is increased at a certain rate with the driver. In this case, the problem can be solved by increasing the flow rate at the same rate, but the energy consumption will increase as the cubic of the increase in the flow rate and the energy consumption will reach very high values. This problem can only be solved by keeping the pressure loss in the jet filter bags at a constant value of around 10mbar.
3.2. Energy Consumption of JPF
JPFs are the equipment that consumes the most energy in the mill after roller mills and pneumatic fans. This energy is consumed during the production of compressed air to be used to clean the filter bags. The power required for compressed air production is calculated by Formula (1).
Here; P is the consumed power (W), μ is the efficiency of the system (isothermal), p1 is the absolute pressure of the sucked air (bar), q is the flow rate of the sucked air (m3/h), and p2 is the absolute pressure of the air leaving the blower (bar). Formula (1) shows that the power required for compressed air production depends on the flow and pressure. In order to work at constant tonnage, the required flow must be produced at a constant rate, and the pressure must be kept at the lowest possible value. In Figure 2, the pressure and power curve of a blower used in compressed air production is given [1]. As it can be seen in the figure, considering that the produced flow rate remains approximately constant, it is seen that the energy consumption moves in the same way as the pressure. This graph shows how effective the pressure used is on energy consumption. Today's energy saving methods in compressed air production always keep the pressure value constant at the usage value, and increase or decrease the flow according to the need. In this case, only the flow can be controlled and the pressure value cannot be interfered with.

It could be seen from Figure 2, that a more successful result can be obtained than today's known energy-saving method if the production and consumption of the compressed air used for cleaning the jet filter bags are synchronized [1]. Thanks to an algorithm to be developed, it is possible to control together both flow rate and pressure, not only the flow rate. Artificial intelligence-based Jetmaster Deep series advanced process controllers reinforced with deep learning have such an algorithm.
3.3. Filter Timer
The operating logic of the devices that control the JPF bag cleaning process, called the filter brain in the industry, remained as it was first developed by Bühler in 1979. Despite the intervening more than 40 years, jet filters are still controlled depending on time. Two parameters are used here. One determines waiting time (off time) and the other determines blow duration (on time). But, it is a completely outdated control method. Even if the pressure is blown at a full 0.5 bar value at the beginning, then by the time, this value cannot be achieved in the normal time due to the ambient and working conditions, clogging of blower filters, humidity, and temperature. For this reason, the time is set longer than it should be. In this case, blowing cannot be done immediately after the pressure becomes 0.5 bar, air continues to be produced. However, this produced air is exhausted through a safety relief valve to keep the tank pressure constant at 0.5 bar. In this case, air that will never be used is produced with the highest power consumption and thrown out by a relief valve. Existing timer controllers are very primitive and are the biggest cause of inefficiency in existing systems. The blower energy consumption that will occur if these devices are used is given in Figure 4.
Figure 2. Effect of Pressure on Energy Consumption
Instead of these time-referenced devices, much more advanced devices with artificial intelligence technology, such as Jetmaster, that can offer algorithms in which pressure and differential pressure combinations are used together, should be used. With these devices, blowing can be done as soon as the pressure reaches 0.5 bar. In this way, inefficiency is completely eliminated. In addition, the pressure can be set to the desired value according to the need. The blower energy consumption to be realized with artificial intelligence devices is given in Figure 4.
3.4. Cleaning Multiple Filters with One Blower
In facilities where more than one filter is cleaned with one blower, the timer of each filter blows at different times. In this case, the filter tanks become pressure vessels connected in parallel and some filter bags are blown at a lower pressure as in Figure 5. If this event is repeated one after another, a permanent cake layer is formed in the filter bags, then the bags lose their permeability. This problem both reduces aspiration and causes the bags to burst when air is blown into the bags. Flour production capacity and extraction rate decrease, fluctuations occur in ash value.
Figure 8. Change of Oil Life According to Operating Temperature
Generally, a blower with a capacity of 2-3 times larger than normal is used to eliminate this problem, that is, to keep the pressure between 400-500mbar continuously. In this case, the pressure is constantly in the range of 400-500mbar as in Figure 6. This ensures that the blowing pressure is high enough. However, the average pressure value almost approaches the blowing pressure value, that is, around 500mbar. In this case, the energy consumption increases between 3 and 5 times. Blower body temperatures, which should normally be around 650C, rise to 1000C or even higher. This high temperature reduces blower life, increases mechanical wear, and spoils the product recovered from the filter.
Figure 3. Energy Consumption When Using Classical Filter Timer
With the JETMASTER Synchronous series devices with artificial intelligence, when the pressure reaches 500mbar, all filters are blown at the same time. By blowing at the precisely adjusted pressure value, very good cleaning of the filter bags is ensured. After blowing, the tank is completely emptied and the pressure is reduced to 0-50 mbar as in Figure 7. Thus, the average pressure value is reduced to around 250-300 mbar. The average pressure value affects the energy consumption logarithmically and the operating temperature exponentially. In this way, energy consumption and heating are reduced by around 85%. Capacity increases by around 10%, ash value decreases by about 5 points, extraction rate increases by around 1%. The same job can be done by a 40% low-capacity blower. Filter bag bursts are prevented.
3.5. Breakdowns of JPF
Failures and malfunctions in the blowers and compressors that supply compressed air to the JPF cleaning system are caused by heating, wear, and friction. The oil used in the blower must be changed at certain intervals. The most important factor affecting oil life is isothermal heating due to the compression of the air. The operating temperature of the compressed air generator is calculated by Formula (2).
Here; T2 is the absolute temperature of the outlet air (K), T1 is the absolute temperature of the inlet air (K), p1 is the absolute pressure of the intake air (bar), p2 is the absolute pressure of the air leaving the blower (bar), and k is the specific air ratio, and its value is approximately 1.4. Formula (2) shows that the heating in the blower depends only on the average outlet pressure. The curve showing the oil life according to the operating temperature of the air generator is given in Figure 3.
3.6. Failures of JPF
JPF is in a structure that will cause many malfunctions that are difficult to detect and diagnose. Some of these glitches are listed below. Models with solenoid valves frequently burn out valve coils. Blowing performance is reduced in models with diaphragm due to diaphragm staling. In single diaphragm models, the tank pressure does not increase when the diaphragm bursts. In double diaphragm models, when the small diaphragm bursts, the tank pressure does not rise, when the large diaphragm bursts, the tank pressure increases, but compressed air cannot be blown into the bags. It is impossible to understand the dirty level of the filter bags. Even if the bags appear clean, their permeability is reduced by moisture. The bag tear cannot be detected until it is seen that dust is thrown out. It cannot be detected that one of the bags has fallen until it is seen that dust has been thrown out. The time to change the bag cannot be accurately determined. It is not possible to always blow into the filter bags at the same pressure. It is impossible to detect the problem of not being able to blow the bags originating from the pneumatic equipment, clogged blower filters, delays in tank filling time, and air leaks in the filter tank. It could not be determined in which filter bag group the fault occurred.

3.7. Type of JPF
According to the pressure of the air used in cleaning the filter bags, those using 6 bar air are called high pressure, those using 0.5 bar air are called low-pressure jet filters. 6 bar air is obtained by using a compressor and 0.5 bar air is obtained by using a blower. Although the investment in high pressure (pulse) filter is cheaper, using a compressor is 2 times more expensive in terms of unit compressed air costs. For this reason, low-pressure jet filters using blowers should be preferred in terms of operating costs.
Figure 4. Energy Consumption When Using Artificial Intelligence Device
3.8. Using the Aspirator Fan Before or After Jet Filter
Aspirator fans can be used before or after the jet filter. In Figure 1, the fan was used before the filter and pressurized the filter. In systems where the fan blows air into the filter, dust particles in the air collide with the fan blades, corrode these blades, and cause balance problems by sticking dust to the blades due to temperature and humidity. The bearings of the fans that work in balance cause breakdowns by disintegrating in a short time. In addition, due to the high air velocity at the fan outlet, dust particles hit the bags much faster, causing the bags to wear faster and explode. Due to the increased filter media pressure, the air density increases, the dust particles become easier to hang in the air, and the separation of dust and air becomes difficult with the cyclone effect. All the load remains in the filter bags. In addition, since there is a positive pressure in the filter, dust will come out from the smallest holes in the filter, which will negatively affect the cleaning.

If the fan is used after the filter since the dust will be kept in the filter, it will not reach the fan and a smoother fan operation will be achieved. In addition, since suction, that is, the vacuum will occur in the filter, and the air density will decrease in the jet filter environment, it will be difficult for the dust to remain in the air. Thus, the separation of dust and air due to gravity will occur in the filter cone to a large extent and less dust load will fall on the filter bags. Moreover, if a hole occurs in the filters, dusty air will not come out from here, on the contrary, fresh air will be sucked in. Thus, the environment will not be dusty. However, if the fan is placed after the filter, the energy requirement will increase by around 2%. Today's modern production approach is not for the fan to pressurize the filters, but for the suction from the filters.

4. Jetmaster: The Industry's First Jet Filterm Timer Using Artificial Intelligence
Today, artificial intelligence has entered into industrial use in many areas. An example of these uses is the jet filter process control device developed by Jetmaster Technology Company (www.jetmaster.com.tr). This device, which has advanced control algorithms with artificial intelligence, is equipped with pressure and differential pressure sensors. The operating method of the device is given in Figure 9. Thanks to this technology, the device offers solutions to all the above-mentioned problems regarding jet filters. The gains obtained according to the data obtained from the facilities where the device is used were as follows.

The production tonnage was increased by 5% and remains constant at the value at which the plant was commissioned. The extraction rate was increased around 1% and reached its maximum value. The ash value was decreased by about 0.05%. Blower energy consumption was reduced by around 45%. Blower production flow was increased by around 20%. The blower oil temperature was decreased around 30°C. The oil change period of the blower was increased approximately 8 times. The mechanical wear on blower parts was reduced by approximately 70%. Waste oil production from blowers was reduced by 85%. The noise level caused by the blower sound was reduced by around 5dB. Blower suction filter blockages and blower failures were detected. The filter bag contamination level was detected. The filter bag burst was detected. Falling filter bags were detected. Filter bag life was extended. Filter bag bursts were reduced. Homogeneous and balanced aspiration was provided at a constant flow. Blower failures and the need for maintenance were reduced. A blower with a much smaller capacity was sufficient to do the same job. The blower investment cost was reduced. If air cannot be blown into any of the filter bag assemblies, the fault is detected immediately. Diaphragm tears were detected. The values given were the average of the data received from the facilities where the device was used. The improvements to be achieved would vary according to the existing inefficiency value. In Figure 10, the comparison of the working methods of the classical filter timers and the artificial intelligence device was given [1]. Figure 11 shows a Jetmaster device in operation.
Figure 6. Energy Consumption and Cleaning Pressure When Using Higher Capacity Blower
[1] Çankaya, N., & Özcan, M. (2019). Performance optimization and improvement of dust laden air by dynamic control method for jet pulsed filters. Advanced Powder Technology, 30(7), 1366-1377.
About the author
Dr. Nihat ÇANKAYA was born in Konya in 1974. He completed his undergraduate, graduate, and doctoral studies in Electrical and Electronics Engineering. He did his master's work on jet filters and his doctorate on control theory.
As the 3rd generation representative of the milling profession, Dr. Nihat ÇANKAYA worked in various positions in factories producing flour, semolina, and pasta, and took part in various projects in important companies of the sector in Italy, Switzerland, France, and Germany. After working as Technical Manager of Selva Pasta, he still works as a lecturer at Necmettin Erbakan University MMYO Milling Program. He is also the Energy Management Coordinator of Necmettin Erbakan University. He speaks English and Italian, and his studies are concentrated in the following areas.
· Project design, establishment, automation, and commissioning of pasta factories and flour mills,
· Energy-saving practices in mills and pasta factories,
· Pneumatic transport systems design and pneumatic calculations,
· High-efficiency jet filter control systems,
· Use of artificial intelligence in flour semolina pasta production,
· Unmanned roller mill systems,
· Increasing the tonnage and extraction rate in flour semolina production, decreasing the ash content,
· Production of oil-free, spotless, crack-free high-quality pasta,
· HACCP, ISO, OHSAS, GMP activities,
· Co-generation facilities design, commissioning, adaptation to food processes,
· Mill and process automation, control systems, PLC, and SCADA applications.
He is also the coordinator of various TUBITAK and R&D projects on related subjects. Dr. Nihat Çankaya has many academic studies, articles within the scope of SCI, and various patents. In addition to his academic duties, he provides sector-specific solutions by designing and manufacturing high-tech process control devices with his Jetmaster Technology Company.