Measuring the quality indices of a charge for coking

The principal quality indices of a charge for coking, as is well known, are its ash, bulk density, moisture and sulfur content. The present designs of ash and moisture meters permit one to determine the ash and moisture of a charge in the flow at any moment of time directly on the conveyor belt or to average the measured values of these parameters for a given time. However, at variable conveyor capacity these values may differ "significantly from the true Weighted mean values. In addition, the quantity of charge passed along the conveyor is not taken into consideration: the charge density is not measured.

On the basis of experimental investigations, the authors propose simultaneous measurement of the ash, density and quantity of charge passed along the conveyor independently of the changes in conveyor capacity, by means of a special instrument. On the functional diagram (Pig. 1), to better show the design of the instrument, the radiation source, detector and conveyor are shown in section along the longitudinal axis of the conveyor.

By means of conveyor belt 1, resting on roller supports 2, the charge enters the measurement zone (the direction of movement of the charge is shown by the solid line with the double arrow). In order to increase the accuracy of meas-surement of the ash and density and simultaneously determine the thickness of the charge layer its surface is leveled and smoothed out by means of leveler 3> made of an elastic sheet material (for example, of strips of conveyor belt). The thick end of the leveler is fixed above the belt, and the thin end hangs freely, leveling and smoothing the charge as the conveyor belt moves. It is also pressed to the flow by means of spring-loaded bar 4 with foot 3» which may-move back and forth in the vertical direction in a guide. The vertical movements are sensed by a type DPK-01 differential transformer sensor. During separation of the conveyor the thickness of the charge layer on the belt may change. This leads to corresponding movements of the bar, which cause changes in the mutual inductance of the sensor windings. The sensor power supply is from generator 7 at a frequency of 20 kHz. Thus, the voltage at the sensor output will unambiguously be determined by the thickness of the charge layer on the conveyor belt. The smoothed and compacted layer of charge then enters the measurement zone.

Under the conveyor belt at a distance of h =5-10 cm in lead collimator 2 in container 8 is 241Am gamma radiation source 9 (gamma radiation energy 60 keV,half life 458 years).

Fig. 1. Structural diagram of instrument for mea­surement of charge quality Indices on a conveyor belt.

 

From it a pyramidal beam of gamma quanta (with an angle β at the apex along the axis of the conveyor and an angle transversely at the apex of β1 < β) falls on the layer of charge through the belt. Above the conveyor belt on the axis of the, beam of gamma quanta at a distance 1, exceeding the maximum charge depth d  by 7-10 cm, is a detector, consisting of an Nalscintillation crystal (Tl) 10, 40 x 40 mm in size, equipped with a type PEU-93 photomultiplier and a preamplifier - pulse former with a photomultiplier. (The photomultiplier and preamplifier are behind the crystal and are not shown in the figure).

The detector is situated in the horizontal plane perpendicularly to the flow of charge so that the center of the crystal lies on the axis of the beam of gamma quanta. The crystal is inside cylinder 11. (On the side of the observer in the figure the bottom of the cylinder is covered with lead, while on the side of the photomultiplier it is open and is about 20 mm above the photomultiplier). In the cylinder there are two diametrically opposed vanes of lead 12 with a central angle a, chosen in accord with the condition

 

                                                                                                                                               (1)

 

where r is the radius of the crystal and R is the external radius of the cylinder.

With condition (1) met the size of the vanes is minimal so that direct gamma quanta from the source do not strike the crystal (in the cylinder position shown in Fig. 1), while the number of forward scattered gamma quanta striking the crystal is maximal. The detector is immovably fixed, and the cylinder may rotate around the crystal due to low-power stepped electric motor 13, controlled by pulses from switch 14. The input of this switch is connected to the output of the photomultiplier output, and the switch outputs are connected to the inputs of integrators 15 and 16.

Under the conveyor belt on a spring-loaded hinged cantilever arm there is' a rubber-covered roller. On its axle there is tachogenerator 17, the output of which is connected to the input of counter 18. The other inputs of the counter unit are connected to the outputs of integrators 15 and 16 and the output of sensor 6. The signals of the computer unit go to indicator 19.

Each measurement cycle consists of several subcycles.

The operation of the instrument is based on the density of flow of the gamma quanta striking the scintillation crystal as a function of the properties and thickness of the layer of charge at various positions of the cylinder vanes relative to this flow. If the cylinder is fixed in the position shown on Fig. 1, then the lower lead vane almost completely absorbs the direct flow of gamma quanta from'the source to the crystal (for this purpose its thickness should be not less than 2 mm). Only the gamma quanta scattered forward in the charge strike the crystal.

The experimental function of the pulse counting rate I from the detector at h = 7 and l = 35 cm is characteristic in nature (Fig. 2): on an increase in the thickness of the layer from 5 to 7.15 cm the signal I increases, and on a further increase in d the signal decreases. The maximum Is observed at =1(where  is the mass coefficient of attenuation of the gamma quanta by the charge p cm2/g). The curve I = f(d) is well approximated by the equation

                                                                                                                                                                        (2)

where Ê is a factor depending on the geometry of the measurements and the activity of the radiation source.

Fig.2. Intensity of forward scattered 1 and passed I1, gamma irradiation as a function of the charge layer depth.

 

The curve was obtained at =100°;=67°; R=24 mm vane thickness 3 mm gap between the crystal and the Internal diameter of the cylinder 1 mm); = 107°22.This ratio of the geometric parameters was taken from condition (1), which In turn provides the minimum statistical error at any assigned time T of a single measurement cycle. In order to obtain the function I = f(d) we used the coking charge of the Dneprodzerzhinsk Coke Works, below 15 mm size class, with bulk density of 0.7 g/cm and ash of 7-5%. At d = 20 cm the sensitivity to density Sp and thickness Sd, are equal and comprise 1.8% decrease of the signal I per 1% Increase in the desnity or layer thickness.

            In this cylinder position the signal I as a function of the ash (Pig. 3)takes the form of. a convex, descending curve. The sensitivity of the signal to the ash at p = 0.7 g/cm, d = 20 cm and A = 10% comprises 1.95% decrease of the signal per 1% Increase In the ash.

After the input of the stepped electric motor receives from the switch a string of 6 pulses, the motor turns and rotates the cylinder by 90°, opening the path for a direct flow of gamma quanta from the source to the crystal; the path of the forward scattered gamma quanta is blocked by the lead vanes, which are now positioned opposite each other in the horizontal direction around the crystal. The experimental dependence of the rate of counting of the pulses from the detector I, in this cylinder position is close to exponential and is well approximated by the equation

 

                                                                                                                                                                       (3)

 

where K - is the geometric factor. In the description of the measurement geometry, 20 cm, the sensitivity to density Sp and thickness Sd, are equal to a 2.9% decrease In the signal per 1% increase in  or d.

Fig. 3. Intensity of forward scattered I and passed I1 gamma irradiation as a function of the charge ash content.

            After the installation of the instrument on the conveyor, it is calibrated before use. The calibration consists of establishing an unambiguous relationship: voltage U, at the output of sensor 6 (see Pig. 1) as a function of the thickness d; signal U2 at the output of tachogenerator 17 and signals I and I1.at the detector output as a function of the thickness d1, , density p and charge ash content A .

After the instrument is cut on a certain number of pulses flow from the switch to the motor, which cause the cylinder to reach the position shown in the figure. At the end of a certain period of time, required for final positioning of the cylinder (t of about 0.05 sec), the switch transmits a control signal to actuate the counting unit and connect the output of the detector with the input of Integrator 15- The first subcycle of measurement with duration t1. begins

During this time the Integrator, with time constant , accumulates the signal I, and the counting unit Integrates the signals

                                                                                                                                (4)

                                                                                                                                                               (5)

 

In order to suppress fluctuations of the signal at the detector output, the integrator time constant is chosen at =5 sec, and t1= 100 sec. At the end of the first measurement subcycle the switch transmits a control signal for temporary cutoff of unit 18 (see Pig. 1) and transmits to the motor a string of 6 pulses, which causes the cylinder to rotate by 90°. After a time of t=0.05 sec the switch transmits a trigger signal to the unit and connects the detector ouput to integrator 16. The second measurement subcycle with duration t2 begins.

During this time the integrator with the constant  accumulates the signal I1, and counting unit 18 integrates the signals:

                                                                                                                                     (6)

                                                                                                                                                      (7)

The time constant of the integrator is, and the time of the second subcycle is chosen from the condition t2=It=36 sec, when the statistical errors of measurement of the signals are approximately equal.

At the end of the second subcycle a signal from switch 14 triggers the counting unit to count the measurement results. Simultaneously, the motor again turns d the cylinder 90°. The values of ρ1 A1d, and m1, are memorized in the counting unit memory. The next two measurement subcycles proceed similarly and the second values ρ2 A2d, and m2 are memorized.

A full measurement cycle consists of 6 pairs of subcycles, at the end of which one determines the weighted mean values of the density, ash and weight of the charge passed along the conveyor belt. The results are indicated on readout panel 19.

At the end of the time t' = 0.05 sec, necessary for the next turn of the cylinder by 90° and to compute the above values, the next pair of measurement subcycles begins. Before this the weighted mean values of the density and ash of the mass are erased from the memory of the counting units, and in the next two subcycles the new values of these parameters are computed. At the end of the seventh pair of measurement subcycles the counting unit also determines the corresponding values, which are indicated on the readout panel.

With this instrument operating algorithms, every (t1+t2)=136 sec the mean values are indicated for the past 13.6 minutes from the measurement of the ash, density and weight of the charge passed along the conveyor.  The full measurement time T = 13.6 minutes, and the time between the replacement of the following results on the indicator (t1+t2)=136 seconds, were chosen as a result of an investigation of the dynamics of variation in the ash and moisture of the charge.  This ratio of T, (t1+t2)and  permits, on the one hand, effective suppression of the effect of random disturbances on the measurement results, and on the other hand, it permits one to reproduce the changes in the measured values with sufficient accuracy. The counting unit may be based on an appropriate micro calculator of microcomputer or constructed in accord, with the operating algorithm from integrated circuits of the appropriate series.  In the calibration of the instrument one determines the coefficients Ê, b, b1, K2, b2, b3, which result in an unambiguous-agreement of the instrument signals I, U, I1, U1 with the ash, density and weight of the charge passing along the conveyor.  The readings are practically independent of the conveyor operation and the lump size and moisture of the charge.  The error of determination of density and ash in the laboratory investigations did not exceed 0.1 abs.% and 0.01 g/cm , respectively.  The instrument is easily constructed from serially produced industrially available parts.

 

REFERENCES

1.  A. M. Onlshchenko and P. I. Grabov, Koks i Khimiya [Coke and Chemistry USSR], no. 10, pp. 7-15, 1979.

2.  N. F. Simonov and A. A. Grigorovich, Koks i Khimiya [Coke and Chemistry USSR], no. 6, pp. 40-41, 1981.