CROSS-CORRELATION INDEX

OF NOISE

IN BURST-PRONE

COAL

AND

NOMINAL

STRENTH

SEAMS

M. S. Antsyferov, N . O. A n t s y f e r o v a , and T. I. Santalova

UDC 622 : 83 : 550.834

To d e t e r m i n e the tendency of a coal seam to bunts or other d y n a m i c phenomena - caving, sudden spills of coal, shock bumps - u s e is m a d e of its seismic a c t i v i t y (noise) [1-4] and the nominal strength index, determined from the depth of p e n e ~ a t i o n of a pointed object into the seam under the effect of a standard force [5, 6]. It is assumed that the noise, i.e., the number of fissures arising per unit time during working of the seam reflects the c a p a c i t y of the seam to undergo brittle fracture. The greater the seismic a c t i v i t y the more readily will the seam fracture and the sooner will i t lose stability. Therefore periods of high seismic activity i n d i c a t e a heightening of the danger of d y n a m i c phenomena. Although very primitive, such a viewpoint is useful because it agrees closely with practical experience: d y n a m i c phenomena are a c t u a l l y manifested during higher seismic activity; this enables one to m a k e a practical seismic prediction of the burst proneness of a seam from variations of the noise [1-4]. As regards the second p a r a m e t e r (the nominal strength index), it is assumed that it reflects m a i n l y the natural strength of the coal seam: the greater the depth of penetration of the point, the weaker the seam and the more likelihood o f the occurrence o f d y n a m i c phenomena [5]. This parameter is used for making local forecasts of the likelihood o f a burst, the a i m being to d e t e r m i n e in good t i m e the tendency of coal seams or their individual sectors to exhibit d y n a m i c phenomena [6]. It was of interest to d e t e r m i n e to what extent both these parameters, which c h a r a c t e r i z e from different angles the strength of a coal seam and its tendency to fracture, depend on rock pressure and to what extent they are mutually correlated. In 1967 a great number of seismic measurements were m a d e by the procedure described in [4] in the steep burst-prone Devyatka seam (sector No. 78, Yunkom colliery, 716 level), accompanied s i m u l t a neously by strength measurements with an instrument designed by G. N. Feit [5]. Furthermore, in this sector episodic observations of the pressure curve were m a d e using a hydraulic transducer designed by L. 13. Famin [7]. In the 10 months of these observations the face advance was 400 m. The strength determinations were performed in the lower part of the w o r k i n g - i n the cross-hole of the haulage road and in the first and second benches. * The measurement lines were located in the gate ends of the face, in the m i d d l e part, and in the foot of the corresponding benches. The measurements were m a d e once per day: in the first and second benches the measurement sites were located at 2 m intervals along the strike of the seam. In the cross-hole of the haulage road the measurements were less regular owing to the nonuniform advance of the latter. Measurements were performed at the surface of the working face; at each measurement point we made five m e a surements of the depth of penetration and then found the m e a n value. This Value, l, is one of the correlation parameters. The seismic measurements were m a d e [1-4] with a geophone, located in the foot of the first bench. The a m p l i f i c a t i o n was kept at such a level that information on cracking phenomena was collected from the first bench and the cross-hole of the h a u l a g e road. A daily count was m a d e of the m e a n number of natural seismic pulses (fissures) per hour of working t i m e of the pneumatic coal pick, n. To obtain the value of nt, the total number o f cracking pulses recorded during extraction in the corresponding 24-h periods was divided by the number of hours of *The Devyatka seam was worked by longwall overhand stoping and the coal was cut by p n e u m a t i c coal picks. The rate of advance of the working front was 2 m per day. A. A.Skochinskii Institute of Mining, Moscow. Translated from F i z i k o - T e k h n i c h e s k i e Problemy Razrabotki Poleznykh Iskopaemykh, No. 6, pp. 9-14, N o v e m b e r - D e c e m b e r , 1971. Original a r t i c l e submitted August 5, 1970.

9 1972 C o n s u l t a n t s Bureau, a division o f Plenum P u b l i s h i n g Corporation, 227 West 17th Street, New York, N. Y. 10011. A l l rights reserved. This article cannot be reproduced for any purpose w h a t s o e v e r without p e r m i s s i o n o f the publisher. A copy o f this article is available from the p u b l i s h e r for $15.00.

616

2

80

January 1967

IFebmar~

March

I April

I

May

I

June

I

Fig. 1. Mean noise curve of the burst-prone Devyatka seam. T h e arrows denote the times of sudden roof subsidence (1) and sudden spills (2 and 3). operation of the p n e u m a t i c coal pick. Thus the statistical relation between the working t i m e of the pneumatic pick and the corresponding number of fissures was assumed to be directly proportional; a preliminary check of this was m a d e in the same seam. The correlation functions of the mean noise n and the m e a n depth of penetration Z were calculated using a computer.* T h e accuracy of assessment of the correlation function was a p p r o x i m a t e l y 20%.

I04

Figure formation in a coal seam being worked is initiated by the v a r i a b l e pressure field. Fissure formation must be most intense in the region of the m a x i ~O,4 mum of the rock pressure curve. This is c o m p l e t e l y confirmed by investigations of the spatial distribution of the seats of fissure formation. It is assumed, of , , J/ ;.~,2 course, that secondary fissuring, arising in the region of the curve m a x i m u m and partly detected by the seismic equipment, must reduce the strength of the seam 0,4: -and therefore increase the depth of penetration of the point. Thus in the zone of the seam corresponding to the m a x i m u m of the rock pressure curve, we must observe a positive e x t r e m a l of the cross-correlation between noise and the depth - 1 2. . - 8 . .- 4 . " 12 of penetration. In the region nearer to the surface of the working face, where the 0,2 ~.'" pressure is less, this correlation decreases. It may evidently fall to zero or b e c o m e negative if the coal undergoes elastic compression here, a c c o m p a n i e d by quasiFig. 2 strengthening. Thus an assessment of the approximate depth of the m a x i m u m of the rock pressure curve can b e m a d e from the position of the first positive m a x i m u m on the right-hand branch of the cross-correlation between noise and depth of penetration.

~\/

,

o~'8,

-

,

T ~ -

,

,

Investigations and d i r e c t determinations of the coordinates of the seats of fissure formation by the s e i s m o acoustic l o c a t i o n method [2] show that fissure formation takes place not only in the zone of m a x i m u m stress, where shear fractures and c l e a v a g e fractures m a y appear, but also in the region of increase in pressure from the steady l e v e l to the m a x i m u m of the curve, where shear fractures m a y apparently be formed. In the region of pressure increase, the stressed state is most l i k e l y uniform only on average, and the state of hydrostatic compression in this region must be evaluated statistically. The individual nonuniformities o f the coal exhibit stress nonuniformities; not only normal, but also tangential stresses appear and the resultant displacements lead to restoration of the statistical equilibrium of the m e d i u m as a whole. Inasmuch as the brittle fractures may h a v e an irreversible character and the fissures formed (even the shear fractures) may expand i f the pressure at the surface of the working face fails, fissure formation during extraction work must affect the strength of the seam and therefore the depth of penetration o f the point. This hypothesis of brittle fracture of coal suggests that the m a x i m u m of the n o i s e - d e p t h of penetration correlation is correlated with the region of the m a x i m u m of the rock pressure curve, but that c o m p l e t e spatial c o i n c i d e n c e of these m a x i m a may be absent. The second index which was assumed to be d e r i v a b l e from the cross-correlation o f noise and depth o f penetration is the recurrence o f settling of the m a i n roof. The periodic character of this process, which is clearly manifested in gently sloping and steep seams, is not readily ascertained. However, seismic phenomena i n d i c a t e that steep seams in the Central Oonbass (dip 60-70") exhibit periodic increases in noise, which, as postulated in [2], are due t o b r e a k way of the sagging rock span, i.e., subsidence of the main rooL Such an interpretation of the periodic a n o m a l i e s

*The computer was programmed by G. L. Kogan, a graduate of the institute.

617

o f increased noise is m a d e further realistic by the fact that j a m m i n g and b r e a k a g e of the supports and caving of the i m m e d i a t e roof are quite often correlated with them. In the periods preceding settling of the m a i n roof, the pressure on the coal seam increases and fissure formation in the seam is intensified. If variations of the rock pressure affect the noise and depth of penetration, the periodic character of the pressure variations, due to periodic sagging of the rock cantilever, must also be manifested in the cross-correlation of these two parameters [8]. Let us d e a l with the cross-correlation functions obtained for the Devyatka s e a m in 1967. Functions were also calculated for two periods of steady cracking. The first period of low noise was from January through April 1967 (Fig. 1), the average noise being 3.64 pulses/h; the second period, with a dangerously high level o f noise, was from May through June 1967, the average noise was 40.9 pulses/h and two sudden spills occurred. These periods exhibited the s a m e depth of penetration and it may be assumed that in terms of this p a r a m e t e r the property of steadiness was not impaired throughout the observations. Figure 2 shows the cross-correlations obtained for these periods. Curves 1 and 3 denote the cross-correlations of noise and penetration depth in the gate end of the first bench; curves 2 and 4 are the corss-correlations of noise and averaged penetration depth. For averaging we took the values obtained in the gate end and foot of the first bench and in the cross-hole of the haulage, road. The cross-correlation of noise and penetration depth in the gate end of the first bench for the period of low noise (Fig. 2, curve 1) lends itself to the clearest physical interpretation. With zero l a g in time, the two parameters exhibit virtually no correlation. This explains the unsuccessful attempts to find a relation between noise and the nominal strength index, which were recorded on the same dates. In the period of days 1-2 (2-4 m from the face) we observe a slight negative correlation ( p ~ 0.05), which is probably due to the a b o v e - m e n t i o n e d secondary quasistrengthening at low pressures in the zone adjoining the surface of the working face. The first m a x i m u m of the crosscorrelation ( p ~ + 0.3), corresponding to the m a x i m u m of the rock pressure curve, is observed with log in t i m e r =4 days, which corresponds to a depth of 8 m. Unfortunately, during this period no measurements were m a d e of the normal stress curve with a liquid transducer. However, a m a x i m u m dep*h of the order of 8 m is perfectly realistic, because, according to [2, 3], in low-noise periods the curves flatten out and run into the body of the seam to a depth of about 6 m. In addition to the position of the first m a x i m u m , the positive branch o f the cross-correlation clearly exhibits a periodic component with a wavelength of 12-14 m (r ~ 6-7 days). This component apparently corresponds to the caving interval of the m a i n t o o l In one of the seismic m a x i m a , which c a m e within this c y c l e on March 13 1967, we did a c t u a l l y observe a slight caving, most probably due to subsidence of the main t o o l Curve 2 on Fig. 2 shows the cross-correlation of noise and penetration depth, averaged for the lower part of the working face, including the first bench and the cross-hole o f the haulage road. The curve was constructed for the s a m e low-noise period as curve 1. The s a m e principal characteristics of the correlation, observed in the previous case, a r e retained. It is true that the first m a x i m u m on the positive axis, corresponding to the depth of the m a x i m u m of the rock pressure curve, is displaced somewhat to the righc r m a x ~ 5days or 1 0 m . The less distinct character of curve 2 is apparently due to the distorting effect of the nominal strength index of the cross-hole of the haulage road. The point is that the data for the cross-hole are incomplete, the readings being often at 5-6 m intervals, and the m e a n penetration depth was 20% less than in any other measurement tine in the overlying benches. T h e cause o f such deviation is not quite clear: it might be due either to the nature of the strength anomaly or (more probably) to the marked difference b e t w e e n the stressed state in the face of the cross-hole and in the overlying benches owing to the cross-holes l i m i t e d size and advanced position. Curves 1 and 2 exhibit a periodic component, 12-14 m long, corresponding to subsidence of the main t o o l Figure 2 shows the cross-correlation in the dangerous zone o f high noise (curves 3 and 4). As in the previous case, a more regular character is displayed by the n o i s e - p e n e t r a t i o n depth correlation in the gate end of the first bench (curve 2). The principal characteristic of this function in comparison with the first case of low seismic activity is that the first m a x i m u m of the cross-correlation approaches the origin. On curve 3 the h i g h - c o r r e l a t i o n region is observed in the 0-1 day (0-2 m) interval, and on curve 4 in the 0-2 day (0-4 m) interval. The fact that the first m a x i m u m approaches the origin ties in well with the fact that the m a x i m u m of the rock pressure curve closely a p proaces the working face; according to G. K. Boiko [2, 3], in the dangerous zones this m a x i m u m is located at a depth of 1-3 m. Direct measurements by Boiko of the curve in the dangerous zone, for which the use a Famin transducer, showed that the depth of the m a x i m u m was of the order of 4 m. Curve 3 clearly exhibits a periodic c o m ponent with a wavelength of 8 m (4 days), which obviously corresponds to the subsidence interval of the main roof in the dangerous zone of high noise. Curve 4 displays a component of even higher frequency (wavelength 4 m). Thus transition to dangerous high-noise conditions is a c c o m p a n i e d by an increase in the frequency of roof subsidence. This increase in frequency apparently readily passes into conditions of continuous caving of the roof, which corresponds to a decrease in the stability of the coal seam in the dangerous zones.

618

The region of negative values of the correlation with r ~ 2 - 8 days probably corresponds to quasi-strengthening of the coal in the zone in which the pressure increases from the steady state in the undisturbed solid rock to the curve maximum in the sector adjoining the working face. Curves 1-4 (Fig. 2) are located mainly in the region of positive values of the function, although the absolute values of the correlation coefficients are low. We may thus assert the presence of a direct relation between the increase in noise and the increase in the penetration depth, i.e., the decrease in strength. However, this relation cannot be said to be close. The closeness of the relation increases markedly in regions corresponding to the maximum of the rock pressure curve; this confirms the hypothesis that the correlation between both parameters (the noise level and the depth of penetration of a point) is due to secondary fissuration, which arises during working of coal seams. These cases of cross-correlations between noise level and penetration depth therefore enable one to assess approximately the depth of the maximum of the rock pressure curve. The correlation functions lend themselves to simple physical interpretation, which confirms the hypothesis that rock pressure affects both the seismic activity and the nominal strength index, determined from the penetration depth of a point. The effect of rock pressure determines the cross-correlation of these parameters, which reaches its highest value at the maximum of the rock pressure curve. Furthermore, these functions enable one to determine the mean caving interval of the main roof, which is not readily determined from primary measurements. LITERATURE 1. 2. 3. 4. 5. 6. 7. 8.

CITED

M . S . Antsyferov, A.G. Konstantinov, and L.B. Pereverzev, Seismic Investigations in Coal Seams [in Russian], Izd-vo AN SSSR, Moscow (1960). Symposium: The Use of Seismic Methods in Mining [in Russian], Nauka, Moscow (1964). Symposium: Geophysical Investigations in Mining [in Russian], Nedra, Moscow (1969). Provisional Procedure for Predicting the Likelihood of a Burst in a Coal Seam by the Seismic Method [in Russian], Izd. IGD im. A. A. Skochinskogo, Moscow (1968). G . N . Feit, Strength Properties and Stability of Burstprone Coal Seams [in Russian], Nauka, Moscow (1966). Provisional Instructions for Predicting the Likelihood of a Burst in Donbass Coal Seams (Plan) [in Russian], Izd. IGD im. A. A. Skochinskogo, Moscow (1966). L.B. Famin, Technology and Economics of Coal Extraction (Tekhnologiya i ~konomika Ugledobychi) [in Russian], No. 4 (1960). W.B. Davenport and W. L. Root, Introduction to the Theory of Random Signals and Noise [Russian translation], IL, Moscow (1960).

619