Search Documents

The question of earth models has repeatedly arisen in our work. An accurate representation of the seismic properties is required for three reasons: 1.To be able to determine from a set of received arrival times, the location of a seismic source in the earth, and secondly the degree of accuracy required in the model used for calculating source location from the set of arrival time differences. 2. To be able to determine the expected arrival times at an array of geophones produced in response to a hypothetical source in the earth. 3. To be able to estimate received signal strengths more accurately. The work reported here on earth models is based on information from sever- al consultants and from others who expressed an interest in the problem and supplied information to us. II. COAL MINE GEOLOGY On one of our earlier assignments for the Bureau of Mines, Earth Science Research Inc. of Cambridge, Massachusetts prepared for us a short description of the lithology to be expected at coal mine sites. This description follows, together with the ranges of compressional seismic velocities (Figure I*) associated with typical materials comprising the layered structure of a coal mine area. Coal is primarily associated with fresh water sediments including sandstones, shales, and clays, and occasionally may be metamorphased to varying degrees. In the Appalachian Basin the coal beds occur within cyclical sequences of non-marine shales, sands tones, conglomerates, limes tones, and clays. The beds range in thickness from a few inches to 60 feet, averaging between 2 feet and 10 feet. They have generally broad areal extent, occasionally up to 5,000 square miles, and have an overall tabular or lens-like shape. Within a particular lens, local changes in thickness are normal.

Jan 1, 1974

During the first quarter of 1973 ADL worked closely with PMSRC staff, under an extremely compressed time schedule, on all aspects of the detailed planning, organization and overall coordination and preparation of the lectures and in-mine demonstrations for the Bureau's Technology Transfer Seminar on Mine Communications given in March of 1973 to representatives of the mining industry. Our efforts centered on the detailed organization and preparation of each of the five technical seminar lectures and associated supporting material, and the coordination of tasks assigned to ADL and other seminar participants. Revised and expanded drafts of the seminar papers were prepared from rough draft versions obtained from PMSRC, Collins Radio and ADL staff. Measurements were made to gather key data on the operation and performance of each of the following systems installed in the Bruceton Safety Research Mine; call alert paging, roof bolt paging, section wireless radio, and guided wireless radio systems. These data were analyzed, manufacturer's equipment specifications utilized, and calculations made to provide performance curves and data for each of the above systems. A substantial amount of artwork was conceived and prepared for 35 m slides, system identification posters in the mine, and seminar handouts. This artwork included some forty 8-1/2" x 11" line sketches, graphs, block diagrams, circuit diagrams, coverage maps, and other line drawings; five 30" x 40" block diagram system identification posters suitable for in-mine use; a 20" x 30" detail perspective illustration of a representative coal mine layout; forty viewgraphs; and seventy 35 m slides of the artwork and word charts. Conferences, discussions and review sessions were held with conference participants to help coordinate the various efforts, obtain technical information, and make final plans regarding the lecture presentations, associated visual aids, and the communication system installations to be demonstrated in the Bruceton mine. These meetings included final systems checkouts and dry run presentations and reviews of the lectures and mine demonstrations. This phase of the seminar effort concluded with R. Lagace and R. Spencer of ADL participating in two days of seminar presentations in the lecture hall and in the Bruceton mine.

Jan 1, 1974

The instrumentation specification in the National Academy of Engineering (NAE) report of March, 1970, set forth some rather general guidelines that the seismic surface system should follow. The fact that there was no comparable equipment in existence at that time, and the need to produce a system from presently available equipment, was spelled out in that document. The features that the system should include were listed, and included: • Quick and simple deployment • Portability • Battery-powered operation • All-weather operation • Computational simplicity and speed • High reliability and long shelf-life • Dual service for location and communication • Adaptability to widely varying seismic velocities. topography, and geologic structures • Minimum maintenance • Capability of continuous operation for extended periods, and • Reasonable cost Since that report, there has been a substantial amount of field work addressing the trapped miner detection and location problem carried out by Westinghouse Corporation under contract with the Bureau of Mines. Two complete surface systems have been designed, built, and field tested and substantial documentation has been produced on the mine environment and background noise using these systems. These systems, built from off-the-shelf equipment, have been designed with maximum flexibility in mind in order to explore what con¬figuration a field unit should have.

Jan 1, 1974

During the summer of 1973 we were asked to perform a theoretical investigation of the propagation of low frequency (LF) radio waves down deep (10,000 feet) hoistshafts* for the case where the hoist cable ("rope") is the only metal conductor present. Propagation is by means of the TEM coaxial mode of transmission in which the hoist cable serves as the inner conductor and the surrounding rock acts as the outer conductor. Since the rock is a relatively poor electrical conductor the current in the outer conductor of the coaxial line is not confined to a very thin surface layer as in a metal coaxial cable, but spreads radially to a distance that is generally many times the shaft diameter. This feature of the wave propagation requires a more sophisticated theoretical treatment than the approximate skin-depth theory that is adequate for metal coaxial lines. In this Part we treat the hoist shaft wave propagation loss, characteristic impedance, and the field current distributions in the surrounding rock medium. We also show that the large penetration of the wave into the rock outer conductor does not present a difficult problem with regard to coupling the transmitter or receiver to the transmission line with a minimum of insertion loss, but that the large impedance mismatch caused by the capacitance termination between the cage and shaft wall may well be the most significant contribution to overall system loss. Inductive coupling and impedance matching to the hoist rope/shaft transmission line are also treated briefly. Preliminary results indicate that a broad minimum in overall system loss should occur between 100 kHz and 1 MHz, possibly centered around 300 kHz. Further work is needed to better quantify this signal loss behavior, compare it with hoist shaft electromagnetic noise spectral data recently acquired by NBS, and identify the most favorable operating frequencies.

Jan 1, 1974

By Wm. C. Dean

The location of trapped miners from their seismic signals will be in- accurate if we assume the P-wave propagation velocity is a constant. P- wave velocities are anything but constant in regions about mines, so some calibration is necessary to obtain more accurate seismic locations. The reference event method compares the time arrivals of signals generated by the miners with those previously recorded from a reference or calibration event in the vicinity of the miners. This method locates the miners' position relative to the calibration source. Hence, the location of the miners is absolute if the reference event position is known absolutely, usually from surveys and mine maps. Advantages resulting from the system, besides greater accuracy, are locations independent of the velocity model assumed, the same solution from the full array and from any subset of four or more seismometers in the array (three if the miners' depth is known), fewer seismometers required, and no complex computers required for analysis. VELA Uniform experience shows that the accuracy of locations of teleseismic explosions and earthquakes is improved by an order of magnitude over locations computed from average travel time curves. Each calibration event is applicable only over a limited range. We recommend a field test of the method at a mine to measure its location accuracy, the range of effectiveness for each calibration event, the number of seismometers needed, and the number of reference events required per mine. From these experiments we could decide whether the reference event method was useful and, if so, what form a practical rescue system would take.

Jan 1, 1974

Crosson and Peters treat the errors that result in miner location due to errors in the overburden earth model used for computing location. A parallel effort by J. Powell of PMSRC is discussed here, where the location method used is that of the Westinghouse location program - "Miner". This location program was tested by Powell by using arrival times generated from the overburden earth model of Table l* rounded to the nearest millisecond. Location computations were then made using 3 geophones at a time. The average location is tabulated for all geophone triplets except when 3 geophones are in line. Geophone arrays were as shown in Tables 2 and 3. The manner in which the test was done is illustrated in Figure 1. I. SUMMARY OF RESULTS The arrival time differences, based on the Table 1 earth model, together with geophone locations based on the array geometries of Tables 2 and 3, were entered into the computer, together with a stated depth of 700 feet and an estimated overburden seismic velocity of 10,000 feet per second. These parameter values were processed by the location program - "Miner". Figure 2 illustrates the interpretation of the plots and data. The tabular data and plots of Table 4 and Figures 3, 4, and 5 illustrate the location error results obtained. These plots indicate that for sources within the array, the errors are considerably less than the measured errors obtained during field tests of the present location system. The possible reasons for this discrepancy are noted in Part Eight (Earth Models).

Jan 1, 1974

This Part is concerned with the theoretical study of UHF radio communication in coal mines, with particular reference to the rate of loss of signal strength along a tunnel, and from one tunnel to another around a corner. Of prime interest are the nature of the propagation mechanism and the prediction of the radio frequency which propagates with the smallest loss. Our theoretical results are compared with measurements made by Collins Radio Co. This work was conducted as part of the Pittsburgh Mining and Safety Research Center's investigation of new ways to reach and extend two-way communications to the key individuals that are highly mobile within the sections and haulage ways of coal mines.

Jan 1, 1974

Over and above the technical support arid consulting work described in the preceding Parts of this Volume, ADL staff also provided a wide range of additional technical assistance to the Bureau on an ad hoc basis over the course of the contract, and particularly in 1973. Some of these assignments were on tasks that temporarily assumed a high priority at PMSRC, requiring a fast response, while others were more suited to conventional schedules but on communication and location topics less directly related to those in the preceding Parts. Chapter I briefly describes some of the diverse short-term assignments while Chapters I1through V treat assignments on the other topics. I SHORT-TEXM ASSIGNMENTS A sampling of short-term, fast-response assignments undertaken on mine communications and miner location for PMSRC during 1973 are given below. These assignments included technical reviews, discussions and recommendations related to the theory, experimental data, designs, and hardware implementation of experimental CW and pulse miner location EM transmitters; preparation of technical performance specifications for a miniaturized preproduction prototype CW electromagnetic location transmitter; review of specifications for a seismic signal detection experimental study; participation in technical discussions and experiments with PMSRC and equipment suppliers regarding the UHF radio/radiax-cable communications installation in the Bruceton mine; participation in an informal conference to discuss mine communication and monitoring needs and preferences with representatives of the Bureau, mine operators, and equipment manufacturers; participation in technical discussions with representatives of a U.S. mine operator and CERCHAR* concerning the characterization of faults on trolley lines and means of improving their rapid detection;** recommendation of simple signal, noise, and ground fault measurements on mine phone lines; review of an inexpensive automatic telephone dialer; preliminary discussions regarding the feasibility of adapting and assembling a collection of available telephone industry carrier equipment for installation and test in an operating coal mine; evaluation of a proposed call alert mine paging system; review of hardware designs and future experiments for a mine communication sled; design critique of a proposed six-channel narrow-band FM voice modem for carrier communication on the mine phone line; preparation of technical performance

Jan 1, 1974

By Francis Crowley

Elementary concepts of seismic wave properties are reviewed to establish a framework for discussing seismic noise and its suppression. A liberal number of references is included for those who wish to pursue the topic in detail. A base motion noise level is established for a single surface seismometer. It is argued that simple array processing and seismometer burial should regularly permit us to approach this base noise level which is found in areas free of conspicuous man-made seismic noise sources. The sensitivity of the base seismic level to a number of common disturbances is presented. Solutions that promise to hold the "effective" seismic noise at a site near the base level are given for acoustic, meteorological and fixed local surface sources. Guidelines are suggested for controlling other local man-made sources, e.g. vehicular traffic, by keeping them at sufficiently large distances away from the seismometers to keep noise at permissible levels. These guidelines are speculative and conservative. They should be the subject of future study. The ability to process against general local activity, drilling, and intra- mine sources is limited. Certainly no dramatic gain is to be expected. When strong man-made and/or intramine noise sources are not controllable, detection may well be impossible to achieve, unless the seismometer can be moved close enough to the suspected miner position, i.e. via very deep holes, to compensate for the likely vast differences in strength of the miner signal and the uncontrollable man-made noise sources. However, many intramine sources, e.g., falling water, rock bursts, explosions, etc., may well preclude the existence of a hale miner in their vicinity, and therefore any need to attempt detection of a miner at such locations. The following recommendations are made to enhance the detection and location of a miner signaling seismically. They are: 1. Seismometer burial in slim holes. 2. Narrow band detection, using multiple narrow band filters to cover the likely signal band. 3. Broadband recording and analysis for time-of-arrival estimation.

Jan 1, 1974

The first phase of our work for PMSRC on this contract was devoted to a comprehensive and in-depth assessment of electromagnetic (EM) noise measurements and data taken by several contractors for use in the design of operational and emergency mine communications. Pertinent measurements and data of other selected investigators were also included in the evaluation. This assessment treated the measurements and instrumentation used, the data analysis and presentation methods, and the utility of the final results. This phase concluded with the identification of the most useful results and methods, the remaining data gaps, the contractor most qualified to fill these data gaps, and the definition (with PMSRC and this contractor - NBS, National Bureau of Standards) of a follow-on noise measurement program, instrumentation system, and data processing methods tailored to obtain the necessary but missing noise data. A partial bibliography of references related to electromagnetic and seismic noise and propagation in the frequency bands of interest was also compiled during this first phase. The body of this Part of the final report presents a summary of the findings of our EM noise assessment, together with detailed recommendations regarding the new noise measurements to be taken and the instrumentation and data processing methods to be used. This work was done during the latter part of 1971 and the early part of 1972. During subsequent phases of the contract ADL was asked to perform on-going evaluations of the continuing program of electromagnetic (EM) noise and propagation measurements made by Bureau contractors, and to utilize pertinent data from these measurements to make performance estimates for

Jan 1, 1974

This Volume documents the Phase 111 effort on the Seismic Detection and Location of Isolated Miners on Contract H0122026 undertaken during the fall of 1972 by a task team composed of ADL staff and several seismic consultants. The team was assembled specifically to work together on complementary tasks, at an accelerated level of effort for four months, to meet the Bureau of Mines' time schedule for obtaining independent, objective, technical judgments regarding seismic methods and systems for detecting and locating isolated miners. The impetus for this work resulted from a compilation and analysis, during the first half of 1972, of new experimental data obtained from a series of in-mine field tests conducted by Westinghouse Corp .+ using the CMRSS* interim seismic location system. The Bureau of Mines took advantage of the availability of these new data to reassess the potential and limitations of various seismic methods and systems, and to direct its seismic system improvement program accordingly. ADL assisted the Bureau in this reassessment by drawing on the skills of seismic consultants from industry, universities and government, to supplement the skills of the ADL project team. The consultants were principal resources of broadly-based and detailed technical expertise in the areas of seismic signal-source and signal-propagation characteristics, natural and cultural seismic noise, seismic sensors and field instrumentation, seismic signal and data processing for detection and location, and overall seismic system utilization in the field under operational emergency conditions. Specifically, the participating consultants were: F. Crowley, Air Force Cambridge Research Laboratories ; W. Dean, Teledyne Geotech, Alexandria Laboratories; R. Greenfield, Pennsylvania State University; J. Kuo, Columbia University; D. Peters and R. Crosson, University of Washington; and F. Pilotte, U.S. VELA Seismological Center. The principal ADL participants were J. Ginty, R. Lagace, M. Roetter, and R. Spencer.

Jan 1, 1974

This Part describes a mine pager phone to public telephone interconnect sys tem that permits mine paging telephones to be selectively interconnected with the public telephone system, as illustrated by Figure 1. This interconnect concept was first conceived and demonstrated in breadboard form by H. E. Parkinson of Industrial Hazards and Communications, PMSRC, as one means for providing improved emergency and off-hours communications. The third-generation prototype unit described in this Part was designed and fabricated by Arthur D. Little, Inc. This unit is a small desk-top or wall-mounted unit that connects directly to both the public telephone line and the mine pager phone line, and requires no modifications to either the public dial telephone or mine pager phones. As part of this task, ADL also built, under a very compressed time schedule, a second-generation prototype unit suitable for reliable demonstration of the interconnect concept during the Bureau's technology transfer seminar on mine communications in March 1973. * A major objective of the design of the third-generation unit described in this Part was that it be compatible with present mine pager phone systems utilizing DC voltage on the mine phone line to activate the phone loudspeakers. Therefore, this unit was designed to operate in a push- to-talk, push-to-listen mode that accommodates the inherent design differences and operation modes of the mine phones of different manufacturers. Though this push- to- talk, push- to-listen mode of operation may seem slightly unusual at first sight, we have found it is easily learned by new users. However, mine phone manufacturers wishing to expand their product lines with such an interconnect system will probably find it to their advantage to customize the design of this system to their particular mine phones, thereby utilizing the unique attributes offered by their individual product lines. This customization should not be difficult, and Section III below suggests some of the features that can be easily added to future designs. Detailed parts lists and circuit schematics for the present third-generation proto- type system can be obtained from the system manual submitted to PMSRC and placed on public file.

Jan 1, 1974

By John T. Kuo

The miner's seismic source strength can be approximately estimated on the basis of a single force acting in an infinite medium. Because of the uncertain- ty involved in estimating the amount of conversion from mechanical energy to seismic energy, the source strength thus estimated may be in error by as much as a half-order of magnitude. Calculations of wave transmission must take into account geometrical spreading, dissipation of energy by internal friction, and energy partition due to reflection and refraction of waves impinging on inter- faces in the layered earth. Theoretically derived peak-to-peak particle velocities in microinches per second (PIPS) are given for two models ; viz: (i) 50-feet thick and (ii) 100- feet thick 4000 ft/sec layers with Q = 20, overlying a half-space of 10,000 ft/sec material with Q = 5O;for the cases .of a hammer blow and a timber impact, at the frequencies of 50 Hz and 100 Hz. Comparison of the theoretically derived peak-to-peak particle velocity with experimental data taken at the Copper Queen Mine indicates that the theoretical particle velocity may be overestimated. A discussion of the distortion of seismic wavefronts by mine tunnels indicates that it is unfavorable to use a seismic source impact on the floor of the tunnel, since deceptive delays in arrival time are liable to occur at the surface. A program of parallel theoretical and experimental work is required to clarify uncertainties still associated with the nature and strength of the miner's seismic source signal which cannot be resolved within the approximations of this work. Its major components include theoretical investigations of the (1) Wave diffraction and scattering of an impact source on a face of a cylindrical cavity. (2) Impact of an elastic object on an elastic medium.

Jan 1, 1974

By Frank Pilotte

Proper design characteristics for a seismic system to locate miners during a mine disaster are considered in terms of a total system concept. Field hardware, processing equipment, operating personnel, deployment, training, and supportive R and D are necessary elements of the total system. Each of these elements is discussed and an indication given as to the requirements necessitated by each one. Hardware must be portable, rugged, and fieldworthy. Processing equipment must be simple to operate, expandable for future processing improvements, and ruggedized for field use. A small well-trained team must be available for quick response in times of emergency. Alternative deployment procedures are necessary to allow for a variety of situations including both ground and air delivery. Supportive R and D is used to support training and upgrading of the system.

Jan 1, 1974

This part of the final report treats some problems related to coal mine carrier frequency communication systems using the trolley wire/track transmission line. This work was undertaken for a brief period of time during the summer of 1972 to help understand, quantify, and improve some of the trolley wire carrier system behavior observed and/or predicted. Calculations and estimates of transmission line characteristic impedance, inductance and capacitance per unit length, mine motor loading effects, parallel line effects, and connection impedances are presented in Chapter I. Chapter II treats the design of r-f isolators for reducing the undesired and troublesome r-f loading that mine motors present to the trolley transmission line. Chapter III treats a diversity method for combatting standing wave nulls. I. CALCULATIONS RELATED TO TROLLEY WIRE COMMUNICATIONS A. ESTIMATION OF INDUCTANCE AND CAPACITANCE PER UNIT LENGTH OF TROLLEY LINES To estimate the L and C per unit length of trolley lines, we use the eccentric line model of Section G of this Chapter, "Characteristic Imped¬ance Estimates for Mine Trolley Lines." The capacitance per unit length for an eccentric line is given by (1)

Jan 1, 1974

By Roy Greenfield

During the detection phase, several procedures can be used to improve the signal-to-noise ratio (SIN) . • Subarrays of up to 10 seismometers should be used. The apparent failure to obtain improvement in SIN in the Westinghouse subarrav work is probably due to excessive noise in the electronics. The spacings needed for sensors in the subarray must still be established. • The sensors should be buried. Even burial to a few feet will probably give appreciable reduction of noise. A deeper burial will probably further reduce the noise; and, if the seismometers can be put below the weathered layer, the signal level will probably be increased. The reduction in noise by burial will be especially effective during high noise periods due to wind or rain. • Before attempting detection, the waveforms should be bandpass filtered. The bandpass should go from 25 Hz to an upper frequency of 100 to 200 Hz. Further measurements with a low noise system must be done before the signal-to-noise ratio can be determined for the 100 to 200 Hz range. Any narrow band noise should be removed by notch filters. • If noise levels differ among seismometers of a subarray, weights inversely proportional to the noise power should be applied to the seismometers. • Automatic methods of screening data for possible signal arrivals should be developed, but the final determination that the signal on a subarray output appears to be from a miner should be made by an analyst. Detection of "signal" on several subarrays should be required for positive identification. In the location phase, after detection: • If repeated signals are received, they should be coherently added (stacked) to give as good a SIN as possible for use in determining signal arrival times, to be used in location computations. The miner should, if possible, repeat the signal 25 to 50 times.

Jan 1, 1974

Coaxial cable structures form the subject of substantial theoretical and experimental investigations either directly related, or which can be extrapolated, to the communications needs and environments of U.S. coal mines. Communication systems based on "guided" waves, of which "leaky" coaxial cables are one implementation, are attractive candidates for providing a range of communication functions primarily along haulage ways, and possible also within sections up to the working face. (However, there are still strong reservations as to whether a practical coaxial cable sys tem could provide sufficient communications coverage within a section. ) In this Part, we present the results of some theoretical calculations of the performance of leaky coaxial cable communication systems. We also analyze the applicability of the theoretical and experimental results obtained by other researchers to the needs of U.S. coal mines. In particular, this investigation has been focused on the utility of relatively low cost, conventional, flexible, coaxial cable at frequencies covering the LF through HF bands and extending into the lower part of the VHF band, as opposed to relatively high cost, special semi-rigid, coaxial cable* at frequencies in the VHF and UHF bands. There are three major design alternatives for coaxial cable communication systems considered in this chapter: • Base station feeding a coaxial cable with high surface transfer impedance (holes in outer conductor, e.g., braid construction). • Base station feeding a coaxial cable with repeaters (U.K. experiments). • No base station, direct communication via a coaxial cable with periodic radiative structures.

Jan 1, 1974

By David C. Peters, Robert S. Crosson

A method of error analysis has been applied to the location technique of non-linear least squares iterative inversion in order to evaluate the resolving power of several seismic array configurations with various assumed earth models and errors and inaccuracies in arrival times. The results obtained demonstrate that lateral accuracies of location are improved significantly when the depth of the miner is known. Lateral location to within 100 feet appears achievable in many instances. If it is possible to refine earth models significantly beyond what has normally been assumed in this work by the use of on-site data, or if the mine is shallow (300 feet or less), accuracies of-about 30 feet or so may be attainable. Inaccuracies in earth models of about 5% are found to contribute much more heavily to these location inaccuracies than errors of a few milliseconds in picking arrival times; however, arrival time errors of 15-20 ms or above will dominate these model inaccuracies. The expected accuracy of location is found to fall off very rapidly as the miner moves outside the array. The strength of this effect depends markedly upon the geometry of the array configuration and can be reduced by careful design. Also, as the size of the array is increased, expected location errors within the array are not altered much, but continue to match the error associated with the smaller array over a larger area (assuming that all stations can still pick up the miner's signals). Better location accuracy, especially with respect to depth control is achievable in an earth where the velocity is depth-dependent (increasing with depth) than in one which is homogeneous. This is an advantage since the

Jan 1, 1974

In this Part, we have included copies of the flip charts and overhead projector viewgraphs used in the two briefings we conducted as part of this seismic detection and location task. The first one was an overview and problem definition briefing given to the seismic consultants on September 7, 1972. The second was a formal oral presentation of study results to technical staff members of the Pittsburgh Mining and Safety Research Center on November 21, 1972.

Jan 1, 1974

By Roy Greenfield

Estimates are given for the distance from a seismometer at which a miner can probably be detected. The procedure in making these estimates was to first establish the natural noise levels at the output of a surface seismometer for the 25 to 100 Hz frequency passband. The noise levels give the range of values which may be expected in areas with no man-made noise. For each noise level, we give the detection threshold which, when exceeded, indicates that a signal has been received. Based on the signals recorded by Westinghouse, curves are given which show the peak signal amplitude as a function of source- to-receiver distance (slant range). Curves are given for the Westinghouse seismic thumper, a 50-pound timber, and a sledge. For a given type of source, the receive signal strength depends more strongly on slant range than on any other factor. However, there is approximately a five-to-one scatter in the amplitudes. Thus, further study of factors affecting the signal amplitude might allow better estimates to be made for any particular geological setting. Combining the signal amplitude with the detection thresholds for the different noise conditions gives the distances at which a miner should be detected. These are given in Table l* below. [ ] The text also gives the increase in detection ranges which should occur if steps are taken to increase S/N by 10 dB. We feel that 10 dB is a conservative estimate of the improvement possible. During an actual rescue operation, the seismic crew and system should be capable of making on-site estimates of their detection capability, based on measurements of the site noise, and upon the best available estimates of the

Jan 1, 1974