Abstract
The largest mechanical pulp production units are high-consistency refiners. In the high-consistency refiners, in-situ measurements of pulp quality are rare. Temperature profile is one of the measurements that is industrially applicable to estimate pulp quality. It measures the steam temperature in the refining area – but how it can detect deteriorated pulp quality will be shown. This paper analyzed the high-consistency refiner plate temperature and its correlation to pulp quality in a 1.34 MW pilot-scale Sunds RGP-44 variable speed refiner in a variety of operating conditions and particularly at production rates close to maximum refiner capacity. The study showed that there was a narrow band of optimal refining conditions at the maximum refiner production level just before pulp quality drastically decreased. When the refiner was pushed above this optimal point of operation, the plate temperature rapidly increased above normal refining temperature, causing reduction of fiber length and consequently lower pulp strength and quality. Thus, refiner plate temperature measurement was found to be a quick way and an effective tool for preventing low-quality pulp production through overloading the refiner.
Download PDF
Full Article
Maximizing Pulp Output and Quality through Measurement of Plate Gap Temperature in High-consistency Refining
Antti Fredrikson,a,* and Jouni Paltakari b
The largest mechanical pulp production units are high-consistency refiners. In the high-consistency refiners, in-situ measurements of pulp quality are rare. Temperature profile is one of the measurements that is industrially applicable to estimate pulp quality. It measures the steam temperature in the refining area – but how it can detect deteriorated pulp quality will be shown. This paper analyzed the high-consistency refiner plate temperature and its correlation to pulp quality in a 1.34 MW pilot-scale Sunds RGP-44 variable speed refiner in a variety of operating conditions and particularly at production rates close to maximum refiner capacity. The study showed that there was a narrow band of optimal refining conditions at the maximum refiner production level just before pulp quality drastically decreased. When the refiner was pushed above this optimal point of operation, the plate temperature rapidly increased above normal refining temperature, causing reduction of fiber length and consequently lower pulp strength and quality. Thus, refiner plate temperature measurement was found to be a quick way and an effective tool for preventing low-quality pulp production through overloading the refiner.
Keywords: High-consistency refining; Thermomechanical pulping; Temperature measurement; Plate gap; Consistency; Energy efficiency
Contact information: a: A Fredrikson Research & Consulting Ltd., Vähäkuja 2a2, FI-40520 Jyväskylä, Finland; b: Department of Bioproducts and Biosystems, Aalto University, Vuorimiehentie 1, FI-02150, Espoo, Finland; *Corresponding author: antti@afrc.fi
INTRODUCTION
The temperature profile of a high-consistency refiner plate gap in the thermomechanical pulping (TMP) refiner has been shown to be a sensitive measure of the stable production conditions for maximum quality pulps. Because temperature measurement of this kind has been tested over the years in many sizes of refiners and it has shown required durability and stability, the findings and suggestions made in this paper can be implemented on an industrial scale (May et al. 1973; Atack and Stationwala 1975; Härkönen et al. 1997; Mosbye et al. 2001; Eriksen 2002; Eriksson et al. 2002; Lawton 2003; Huhtanen 2004; Sikter et al. 2007; Karlström et al. 2008; Fredrikson et al. 2009; Karlström and Hill 2014, 2018).
The analysis outlined in this paper offers new information of the high-consistency refiner plate gap phenomena. Previously, refining and the plate gap phenomena has been explained based on the fiber perspective (Vehniäinen 2008), flow modelling (Huhtanen 2004), entropy modelling (Eriksson 2005), or physical models (Härkönen et al. 1997).
In one of the latest low-consistency refining publications, the optimal plate gap of refining was explored (Harirforoush et al. 2017). At low-consistency refining, the use of temperature profile measurement is different from high-consistency, as liquid water limits the temperature rise over boiling point. In earlier studies, the temperature rise in the high-consistency refiner plate gap has been recognized as plate clash or fiber pad collapse; i.e., the situation when rotor bars touch the stator bars or the occurrence of steam evacuation problems (Allison et al. 1995; Eriksson and Karlström 2009; Karlström and Eriksson 2014).
In the work of Sikter et al. (2007), plate gap temperature had an effect called “r peak”, which is the radial position of maximum temperature at the plate gap. The radial position of maximum temperature depended on operating parameters of the refiner but only minimally. Those measurements were taken in an industrial-scale refiner with large refining area, large radius, and one rotational speed.
In their studies, Miles and Omholt (2008) reported that poor quality, short fiber coarse material was associated with situations in which the pulp pad was loaded with excessive levels of stress in the refiner plate gap. The stress and density of the fiber pad in the refiner has been shown to have radial profiles (Fredrikson et al. 2017; Fredrikson and Paltakari 2018).
The first evidence of temperature profiles or self-pressurization of the plate gap in high-consistency refiners was gathered by May (1973). In previous studies researchers have assumed equal temperatures of the phases: fiber/wood, water, and steam. Furthermore, full saturation of the liquid and vapor phase was also assumed in accordance with Karlström et al. (2008). These assumptions are vital for the feasibility of temperature measurements for the pressure calculation and further estimations.
Is it possible to control the pulp quality with temperature profile measurement in high-consistency refiners? In this study, temperature profile measurements were performed in the high-consistency pilot TMP refiner with simultaneous pulp sampling and quality evaluations. The objective of this study was to find out the joint optimum of unit energy consumption and pulp quality, simultaneously finding out the methods in avoiding collapse of pulp quality.
EXPERIMENTAL
Materials
The refining trials were conducted in the KCL Pilot plant at Espoo, Finland with a pilot scale RGP-44DD TMP refiner that was manufactured by Sunds Defibrator AB (1991; Valkeakoski, Finland). The refiner, with a 1.34 MW variable speed main motor, was used to refine Norway spruce round (Picea abies) wood chips (UPM-Kymmene Ltd., Jämsänkoski, Finland) and also the secondary stage pulps made of the same wood raw material. This refiner is special because it can be operated both in double disc mode (both rotors are running) or in the single disc mode (only feed side rotor running). As shown in Fig. 1, where the feed side rotor is on the right hand side and the load side disc is on the left side of the refiner. It can also be seen how the chips or pulp can enter the refiner from the top (see red arrow), transfer to the refiner plate gap through the channels in the feed side rotor, and finally enter the refining zone where the refiner segments are attached to the discs. The segment pattern type was Metso’s bi-directional RGP42693 at all refining stages (Fig. 2). The pilot refiner used in the trials did not have separate inner and outer segments like the industrial-sized refiners. The chips enter the refining zone at the radius at 330 mm and pulp is discharged at the outer periphery at the radius at 560 mm. Twelve refiner segments form one refining plate and the conical shape cross-section spacing between the feeding side plate and the loading side plate is called the plate gap.
Fig. 1. RGP44SD/DD refiner cross-section. Red arrows show the path of the chips and pulp to the plate gap. Axially loaded shaft with load side stator disc (a) to control the plate gap of the refiner and feed-side disc rotor with main motor (b).
Fig. 2. RGP 42693 segment pattern. The taper (conical cross-section) of the plate gap is shown with grey bars and the clearance between rotor and stator in y-axis. Plate gap at r = 560 mm is typically 0.3 mm. Note: y- and x-axis are not in the equal scale. Breaker bars at radii from 330 to 430 mm and fine section from 430 to 560 mm.
Methods
Plate gap temperature measurement method in TMP pilot refiner
Plate gap temperature was measured at the stator segments with 18 pieces of K2-type thermocouples equipped with Status Instruments SEM203/TC programmable temperature transmitters (Status Instruments Ltd., Tewkesbury, England). According to standard IEC 60584-1 (2003) the tolerance of K-type thermocouple is 0.0075 °C. The thermocouples were mounted to 2-mm diameter holes at the refiner segments with epoxy (Fig. 3). The heads of thermocouples were left visible, i.e., they were not fully covered with epoxy. The thermocouples were mounted on top of the bars and the bottom of the grooves of a segment with 28-mm radial pitch from radius 330 to 560 mm. Note that thermocouples mounted at positions R332 and R560 are not associated to refining area that was exposed to wear. Plate gap temperature was also measured from the rotor side with one PT-100 sensor mounted at the True Disc Clearance (TDC) sensor at the radius of 495 mm. Temperature readings were recorded simultaneously at all radial positions. Calibration of thermocouples was done by flushing the unpressurized refiner with saturated steam and correcting the individual sensor output readings to be the same. The data acquisition was performed with the KCL-Wedge™ system (Oy Keskuslaboratorio – Centrallaboratorium AB, version 6.0, Espoo, Finland) at 0.5 Hz frequency for the temperature and for all other refiner’s on-line process variables (Kahala 2008).
Fig. 3. a) Mounting of the thermocouples to the top of the bars and the grooves of a refiner segment. Blue circles illustrate the bar top and orange circles represent the groove sensor tips. As seen, the heads of the thermocouples were left visible. b) Temperature sensor pitch along the refiner segment radius was mostly 28 mm. The inner most (R332) and the outer most (R560) temperature sensor were not mounted to the bars, instead they were on the side of the segment.
Fig. 4. Plate gap pulp sampling holes at the stator segment. Fifth hole (without arrow) is the bolt hole
Plate gap pulp sampling system
The pilot refiner, with a refiner plate outer radius of 560 mm, was equipped with a plate gap pulp sampling system. In addition to discharge pulp samples, pulp samples were taken from 20-mm holes in the plate gap from the following radii: 400, 435, 470, and 505 mm. See Fig. 4 for the holes positioning in respect of the segment pattern. The refiner had 12 segments attached at the disc and one of them had holes and tubes for the sampling.
From the pressurized plate gap, the pulp samples travelled through flexible metal tubes out from the refiner to the cyclone. High-pressure steam (500 kPa) was blown periodically through the tubes towards the refiner plate gap to be able to flush the pulp sampling ports and tubes. This procedure kept the sampling ports mostly open and ready for sampling and it did not disturb the pulp flow at the plate gap. See Fig. 5 of the illustration from the outside of refiner for tubes, valve assembly, and the cyclone for the pulp sampling.
Fig. 5. The layout of the plate gap pulp sampling system
Running the Refining Trial
The refiner trials were conducted by setting the refining conditions as stable as possible. Refiner casing temperature (or the pressure) was controlled with fresh steam flow, blow back steam flow, and discharge steam flow control. Typically, the fresh steam control was set to manual and the blow back and discharge control loops were given pressure set points. Production rate was controlled by feeding screw rotational speed and that was not changed during the one set of trial points. Dilution water flow was set manually to dilute the refined pulp to approximately 30% dry matter content. The refiner load was controlled by altering the plate gap. There was no automatic control loop for the plate gap in the pilot refiner but two buttons for the operator — to reduce and to increase — the plate gap. Pulp samples were collected with a 4-L scoop from the atmospheric pulp discharge first to 100-L buckets and, after mixing the collected large sample, the actual samples were packed to 4-L plastic bags for the storage. The 100-L pulp sampling was always performed during the whole trial point time (from 5 min to 15 min) by altering the time interval between two subsequent scooping.
Testing the Effect of Temperature Measurement Location at Refiner Segment
Temperature measurement in the modern TMP refiner is a fairly easy task. In the past, in most of the trials the thermocouples were positioned under a protecting sheath. In this study, the assembly location of the temperature sensor head was assessed by evaluating the measured values from the top of the refiner segment bar and the bottom of the groove (Fig. 3).
Different Rotational Speed Trials with Pilot Refiner
Three distinct primary stage refining trials with various plate gap clearances were performed with rotor rotational speeds of 1200, 1500, and 2200 rpm. The same plate pattern (Fig. 2.) was used in all these trials. The refining casing pressure production rate and dilution water flow were similar in all trials (Table 1).
Table 1. Refining Parameters for Rotational Speed Tests
Chips and Pulp Refining — Primary and Secondary Refining Stage
In the primary stage (with chips) refining test 300 kPa over pressure and 1500 rpm rotational speed were used and in the secondary stage (with pulp) pressure was 150 kPa and rotational speed was 1500 rpm. In the secondary stage refining tests, the fiber flow to the refiner was pulp and the Canadian standard freeness (CSF) levels were lower compared to the primary stage.
While keeping the other refining parameters constant (refiner casing pressure, production rate, and dilution water flow), the loading of the refiner was increased via reducing the plate gap clearance, and the changes in temperature profile, axial thrust, motor load, and rotational speed were monitored. Pulp samples were collected in multiple parts during the 5-min trial points to ensure the representativeness. In the primary stage, trial production rate was approximately 800 kg/h. Plate gap clearance was varied from 0.2 to 1.0 mm to obtain various motor loads. Production rate at the secondary stage refining was 524 kg/h. Plate gap varied from 0.2 mm to 1.3 mm and only a small change was made for the dilution water flow (from 10 to 8 L/min, respectively). The pulp produced at each trial point was analyzed according to ISO standards. Fiber length was measured with a Kajaani FS-300 fiber analyzer (Metso Automation Ltd., Kajaani, Finland) according to the ISO 16065-1 (2001) standard. The accuracy of Kajaani FS fiber analyzers is considered precise as the previous version FS-100 reached a standard error of 0.0016 mm with 95% confidence level (Copur and Makkonen 2007). The CSF was determined according to ISO 5267-2 (2001) standard.
Primary and secondary stage refining tests were performed to show fiber cutting in the presence of temperature peaks. “Fiber cutting” was defined to occur when pulp fiber length differed from the logarithmic CSF and fiber length curve (Figs. 13 and 15). The presence of the “temperature peak” occurred when two adjacent temperature readings had a large difference or the outer periphery temperatures differed from the refiner casing temperature (Fig. 12).
Temperature Peak and Fiber Cutting Evaluated with Plate Gap Pulp Sampling
The phenomena of fiber cutting in the low plate gap periods was observed by taking pulp samples from the plate gap with sampling ports. All results and the explanation of the whole plate gap pulp sampling experiment can be found from Sari Liukkonen’s IMPC 2014 presentation (Liukkonen et al. 2014). In the study, the chip raw material was macerated to see the potential of the pulp fiber length, i.e., see the original fiber length of the raw material. The pulp samples taken from the plate gap were also macerated to find out what the fiber length of the full stream is, not only the already-fibrillated part of the pulp. In Fig. 6, the refiner segment pair is shown. The coarse segment pattern was used in the rotor position and the fine pattern (with sampling holes) was used in the stator position. A rotor speed of 1500 rpm was used in the presented trial.
Fig. 6. Segment patterns used in the trial; (a) fine pattern used in the stator position with the pulp sampling ports and (b) coarse pattern used in the rotor position (the extra hole was for the TDC sensor)
RESULTS AND DISCUSSION
Effect of Temperature Measurement Location at Refiner Segment
In normal operation, the stator bar pattern at the groove is filled with stagnant pulp. It was not known how much this dense pulp pad affects the temperature readings at the pilot refiner. After a series of trials, it was shown that there were no differences in the average values between the bar top and groove at all radial positions. However, the standard deviation of temperature was much higher in the bar top location, which corresponded to faster reaction to changes in ambient temperature (Fig. 7).
Figure 7 shows 10 min sampling of temperature data from two sensors at the radial position of 494 mm. It was observed that the temperature at top of the bar (black line) was more sensitive for rapid changes. Temperature value in the groove of segment pattern (red line) was more stable as the pulp on the stator groove acted as a heat insulator. The bar top surfaces were clear of pulp and consequently reacted faster.
Fig. 7. Temperature at top of bar of the segment (black) or groove (red); the temperature at the groove had smaller variation than the temperature on the top of bar
The standard deviation for temperature at the groove was 0.1 °C, and at the top of the bar it was 0.5 °C. Nevertheless, the average values were the same. In this trial, the variation in the bar top temperature was caused by the oscillation in the pressure regulating valve. The oscillation of the valve opening had same time constant with temperature values. The result is important; as the refining segments wear on the bar tops and the sensors in that location will be destroyed before the end of segment-life time. Temperature sensors in the grooves have more lifetime expectancy.
Effect of Rotational Speed to Temperature Profiles
The plate gap range covered the refining net load from 500 kW to 1200 kW, i.e., from Specific Energy Consumption (SEC) 0.63 MWh/o.d. metric tons to 1.50 MWh/o.d. metric tons. Figures 8 through 10 represent temperature profiles with different loads, i.e., with different plate gap clearances. It can be seen that maximum temperature decreased with reduced load, i.e., with increased plate gap clearance, but the radial location of the maximum temperature peak(s) did not change with fixed rotational speed. Maximum temperature appeared at radii of 410, 430, and 470 mm for the respective rotational speeds of 1200, 1500, and 2200 rpm. It seems obvious that higher rotational speed moves the radial location of maximum temperature outwards. This is important for the segment pattern designers when patterns are optimized for refiners operating at different nominal speeds. The frequency of the electrical grid causes the nominal speed to be different in North America 60 Hz and Europe 50 Hz. Same type of refiners are used in both areas. Therefore refiners have RPM 1800 in North America and 1500 in Europe.
The trials clearly showed that the radial location of the maximum temperature in the plate gap did not change within whole operation range of plate gap clearances when rotational speed was fixed (as shown in Figs. 8 through 10).
It is thought that chips or pulp together with water are the main hindering media of steam flow in the plate gap. It was proposed that the relative radial flow resistance for steam did not change significantly within normal plate gap operation range, which in this refiner was 0.20 to 0.65 mm. This is consistent with the results with industrial-size refiners conducted by Härkönen and Tienvieri (1995).
Fig. 8. Surface map of temperature (°C), radius of refiner segment (mm), and refiner motor load (kW); rotational speed 1200 rpm
Fig. 9. Surface map of temperature (°C), radius of refiner segment (mm), and refiner motor load (kW); rotational speed 1500 rpm
Fig. 10. Surface map of temperature (°C), radius of refiner segment (mm), and refiner motor load (kW); rotational speed 2200 rpm
Chips and Pulp Refining: Primary and Secondary Refining Stage
At the primary stage test, it can be noticed in Fig. 11 that in the low end of plate gap clearance the maximum temperature deviated greatly from linear dependency of the plate gap and axial thrust. Furthermore, in Fig. 12 a large deviation was observed in temperature with 0.28-mm plate gap at the radii of 450 mm and above. The deviation can be observed when comparing the temperature profiles with 0.37-mm and 0.41-mm plate gaps. When comparing the temperature profiles of the normal operation range in Fig. 9, it was clear that the rise in temperature occurred at different radii. Exact values of the trial point setup and pulp properties presented in Figs. 11 to 13 are shown in Table 2.
Table 2. Trial Point Parameters, Their Values, and Pulp Properties of Primary and Secondary Stage Test
