ACARP PROJECT C8021

Data Handling for Gas Drainage Flowmeters

Ian Gray, Paul Clemence

February 2002

ABSTRACT

This report describes the development and final form of an electronic system designed to monitor underground gas flowmeters. The system is designed to power and communicate with up to 255 devices along a twisted pair cable which may have a length of up to 2 km. The flowmeters may be replaced with other transducers so that the system may also be used to monitor extensiometers, convergence rods or a variety of resistance bridge type sensors. Whilst the system is not electronically fast it enables any sensor to be accessed in 1.5 seconds.

Ian Gray

Paul Clemence

February 2002

INDEX

1. BACKGROUND

The technology described in this report has been developed to provide electronics for Sigra's underground gas flowmeters. The purpose of the gas flowmeters is to permit the automatic measurement of gas flow from underground boreholes. The original flowmeters were developed under ACARP project C5030 Development of a Gas Drainage Flow Meter. They have been replaced by alternative designs which are much simpler than the original.

2. FLOWMETERS

The flowmeters can take several forms but are in essence simple devices with variable flow restrictions. The flow restrictors are flaps which open with increasing flow and tend to shut down with decreasing flow rates. They may be thought of as self varying orifice plate flow meters which instead of the usual maximum to minimum flow ratio of about four have instead a ratio of about 100. This is useful because of the wide flow range of gas drainage boreholes. The other feature of these devices is that they may pass particles or water and are self cleansing.

The flowmeters are essentially differential pressure devices in which the flap movement is in response to the differential pressure. For flow to be calculated in any differential pressure device the density of the fluid needs to be known. In the case of a gas the density is a function of the gas type, temperature, pressure and moisture saturation. If the gas type is known and the saturation assumed to be 100% then the only measurements that need be taken to obtain density are those of pressure and temperature. The pressure in the flowmeter may lie above or below atmospheric depending on the state of the gas drainage plumbing system. Realistic pressures lie in the -50 to +50 kPa range and this pressure may be simply monitored by readily available commercial pressure transducers. Etched silicon diaphragm devices have been used for this purpose. The temperature may be monitored by simple solid state temperature sensors.

A form of one of the flowmeters is shown in Figure 1. This uses a hall effect sensor mounted on the flap to detect its position between two magnets. A temperature sensor is mounted on the tube to detect the gas temperature and a port is connected to the pressure transducer to measure the gas pressure. The pressure transducer has in turn another temperature transducer placed upon it. This permits temperature corrections to be made to the pressure transducer values.

In another embodiment the flowmeter uses a flap fitted with a magnet. Two hall effects sensors are used in this case. One above and one below the flap. Each hall effect sensor measures with greater sensitivity the closer it is to the magnet.

In yet another form the flowmeter shown in Figure 2 uses a differential pressure transducer to measure the pressure drop from one side of a bi-wing flap to the other. This form is particularly useful as it enables the system to be used in large diameter pipes where the diameter would make hall effect sensors difficult to operate. This also permits the system to be used with pipes that are magnetically susceptible. The limitation of this system is the lower pressure limit range of differential pressure transducers that are available at reasonable cost and that are capable of taking a wet gas.

These effectively limit the flowmeter to a minimum of a 2 kPa pressure drop at full flow.

The electronics therefore requires either:

1)         Two temperature sensors

                        One hall effect sensor

                        One pressure transducer OR

2)         Two temperature sensors

                        Two hall effect sensors

                        One pressure transducer OR

3)         Two temperature sensors

                        One pressure transducer

                        One differential pressure transducer

One of the benefits of flowmeters containing a pressure transducer is that the pressure information monitors the health of the gas drainage pipeline thus enabling pooling of water, pipeline breakage or other malfunctions to be detected.

Flow from underground boreholes is irregular and frequently interspersed with slugs of water. Sometimes the flow velocity becomes high and particles may become entrained within the flow. The flowmeter is designed to survive both of these events though the apparent measured flow is clearly not correct when a slug of water is passing through. In dealing with slug flow it is necessary to read the flow repeatedly, remove obvious water slug effects in software and to be able to average the gas flow over what may be a minute interval.

3. REQUIREMENTS OF THE ELECTRONICS

The basic requirement of the flowmeter electronics is to be able to log a line of boreholes along the length of a gateroad and to be able to transmit this information to surface for storage and processing. The information need not be gathered at a high speed and logging every few hours would be more than adequate for most cases.        

The electronics needs to be intrinsically safe so that it will not cause an ignition. Furthermore it is desirable that it is robust enough to survive impact.

4. OPTIONS CONSIDERED

4.1 Battery Powered Cable Connected Unit

Initially in the project cable communications systems were considered and a logger was built to use them. It included an LCD and a battery and was intended for local or remote use.

4.2 Short Range Radio Unit

The problems of cable survival became a real issue and a great deal of effort was expended on short range radio (<10 mW power) linked devices. These can be made to work readily in a tunnel with 30 m to 50 m range. To be able to cover a greater range they must talk to each other and a daisy chain of communication needs to be set up so that the controller talks to the first unit which then calls the second unit to get the third and so passes the command along until the desired unit is reached. That unit then turns on the flowmeter and takes a reading which must be transmitted back along the daisy chain. The cost of such a transmission system is twofold. The first is time and the second is battery power.

Each transmission takes about 1 second to perform. Thus to get an answer from one flowmeter would take one second to request, half a second to take a measurement and one second to respond. Or a total of 2.5 seconds. To reach two units takes one second to transmit, 0.1 seconds to relay, one second to re-transmit, 0.5 seconds to take a reading, 1.0 seconds to reply, 0.1 seconds to relay and 1.0 seconds to transmit. This is a total of 4.7 seconds to get a signal from the far unit. To this must be added the time to get the signal from the closer unit, 2.5 seconds, thus reaching a total of 7.2 seconds to get data from two flowmeters.

The times increase quite significantly with more flowmeters. Thus 10 flowmeters would take 4 minutes 13 seconds to read all devices. Thirty flowmeters would take 17 minutes to read and 60 flowmeters would take 67 minutes to read. In addition to the time issue is the fact that each flowmeter must spend a significant amount of time in transmission. This requires battery power.

The battery power in each unit was the final straw that broke the systems back. Not from the viewpoint of battery size but for safety and convenience. No matter how strongly the cases were made the limitation of having a battery powered device in a hazardous area did not seem to be a good idea. The potential for a rock fall or some other accident that might lead to crushing of a battery compartment was not considered prudent. In addition the problems with a need for battery replacement could not be solved readily within reasonable cost.

 

4.3 Twin Wire System without Batteries

The final solution arrived at was to power all units by a robust twin wire communications cable and to strap this wire to the gas drainage pipeline for most of its length so as to avoid damage.

The real breakthrough of the system was reducing the current of the system to a minimum so that up to 255 units could be driven from a single twisted pair cable. This enables the cable to be driven through an intrinsically safe barrier. On one side of the barrier is the controller which is connected to either a modem or the mine data highway (normally via a PLC). On the other side of the intrinsically safe barrier is the twisted pair wire which is connected in a daisy chain to the sensor units. A block diagram of this system is shown in Figure 3.

The wires carry 12 V DC and communications signals. Data transmission is half duplex. The system may be in an idle state or the controller unit or one of the sensor units may be transmitting at any one time. Each unit (including the controller) has a transmitter and receiver section. The transmitters consist of a switched current sink. The current is turned on and off to generate a voltage change on the two wires. This voltage change is detected by all the receivers in the system.

All sensors are reset by powering the twisted pair line with 12V. On receiving power the address decoder on the sensor is powered up. The decoders are very low power devices with 10 micro ampere quiescent power (= 2.6 mA current for 255 sensors). They listen to the twisted pair line for their unique eight bit address number. This number is set by adjusting a series of jumpers internally to some number between 1 and 255. On receiving this information, only the addressed decoder wakes up and powers the sensor microprocessor. Other decoders stop sampling until they receive a new reset power on signal. Only the addressed unit draws power off the line. This power supplies the microprocessor which sequentially switches on the various sensors to a maximum power of 20 mA. The microprocessor then transmits this information back along the line using the sensor's transmitter. 

Data transmission is designed to be between 50 and 300 baud over 2 km. Thus at the lowest baud rate data from the sensors may take slightly more than one second to transmit.

The controller is designed to sit in a non hazardous area and to be mains powered. It communicates with the sensors through an intrinsically safe barrier. Logically all of this equipment resides in an (flameproof) enclosure. It is designed to communicate with the mine's data highway through either an RS232 or RS485 link. Alternatively it may be simply connected to a modem. A block diagram of the controller is shown in Figure 4.

5. SOFTWARE - THE KEY TO THE SYSTEM'S COMMUNICATION AND INTERPRETATION

Different mines have varied data highway hardware and software. Making these communicate with the controller requires specialist software to be written. This need not necessarily be a huge task but it will need to be undertaken on a case by case basis.

The data likely to be produced is significant and some data handling software is required. This may operate at several levels. First to examine the information from each sensor and to decide whether the sensor is working. Breaks in transmission or unlikely readings need to be flagged so that investigation and repair can take place. Secondly the pressure readings need to be examined to see whether the gas drainage range is in good condition. Thirdly the gas flow readings themselves need to be able to be examined to see whether gas drainage is taking place. In its simplest form this takes the form of files of data for each flowmeter. In a more complex form the data from each flowmeter can be integrated with respect of time to arrive at a total flow. This can be used to calculate a material budget for the gas in the block of coal in question. The ultimate use of the data is to run it in parallel with a real time reservoir simulator to arrive at gas pressures and contents within the coal seam.

6. ALTERNATIVE USES OF THE SYSTEM

The sensor units have been kept as general as possible. Only the two temperature transducers are dedicated devices. The other four inputs are separately switchable and are fitted with instrumentation amplifiers. They each are connected to the microprocessor which has a 10 bit analogue to digital converter on board.

Thus the inputs may be hall effect sensors, resistance bridges, potentiometers and probably many other devices. This immediately opens the possibility of this logging system being used for roof extension or convergence monitoring using potentiometers. It also permits the use of various bridge type transducers to measure air flow velocity, water pressure or whatever other kind of measurement is required in the mine. The limitation is that the current draw of the transducers used should be kept below 10 mA.

7. LOCAL COMMUNICATIONS WITH A SENSOR

Unlike earlier developments the sensor units do not have a display. This omission simplifies the electronics and makes the electronics housing much easier to build or purchase. Each of these factors keeps the cost down. This however means that it is not possible to walk up to a flowmeter underground and read the flow or pressure. To facilitate this the devices may be simply interrogated by plugging in a two wire connector. This may be connected to a PC on surface or to another IS device with an LCD underground.

This plug in system permits the sensors to be programmed on surface. They are programmed with the calibration coefficients of the various sensors and with their own individual serial number.

Underground it will be possible to walk up to a flowmeter, plug in a hand held read-out, push a button and obtain information from a sensor. To make this possible the calibration coefficients of each transducer are held within memory on board the sensor.

 

8. STATUS OF THE SYSTEM

In its present demonstration form the system is set up with the controller connected to a modem. The controller is connected via an intrinsically safe barrier to a two wire twisted pair cable of 1.0 sq.mm cross sectional area. The twisted pair is in turn connected to five sensor units. The units can each be addressed and each replies to the controller. The modem is addressed via a telephone line from a computer and data is gathered into the computer.

To make the system a full underground unit the intrinsic safety approval process must be completed for the cable, sensors and the transducers with which they will be used. This aspect of the project has not been completed neither therefore has the mine trial. Notwithstanding this John Hanes and Jon Sleeman of ACARP have advised that the project report should be completed at this stage.

The software to handle the data from the flowmeters involves simple data storage of the raw information from each transducer in the flowmeter. This is combined with a file containing calibration coefficients to calculate flow and pressure information. This is stored chronologically in a file for the particular flowmeter. Trigger points may be set to alarm high or low flows or pressures and the data may be used in combination with a gas content file to work out the gas balance. The basis of this is the simple equation

         GAS INITIALLY IN PLACE-GAS DRAINED=GAS IN PLACE

To undertake this calculation the coal must be divided into blocks which are given an initial gas content. The flow from each borehole needs to be ascribed to a particular block. In its current form this is a simple calculation and the result is stored in a file. The system could readily be extended to provide a graphical output of a mine plan with blocks coloured to correspond to their gas content. The shortfall in this approach is the way in which the blocks are manually divided and the lack of knowledge about flow distribution along the borehole. A real time simulation approach with a numerical reservoir model driven by flow data, pressure information and updated gas contents would in the longer term provide a better basis for interpolation.

The flowmeters can be built to order. The differential pressure device is particularly simple and low cost. The prime limitation currently faced is one of calibration. Currently Sigra can calibrate devices up to 45 litres/second flow (air at STP conditions). Building bigger devices requires a calibration facility of substantial capacity to cope with the requirements of some mines to handle flows of 500

litres/second. This corresponds to a compressor of 1100 cfm capacity. Industrial fans and blowers do not generally have the pressure rating (5 kPa) to permit them to be used for calibration and those that do are large expensive units that require several kilowatts of power to drive them. Whilst a calibration facility can undoubtedly be built to do so it requires a market that will definitely buy flowmeters to justify the cost.

Small flowmeters have been used successfully by Sigra in their own operations. The prime limitation of the use of these underground has been their inability to pass large particles of coal. A flowmeter has the ability to pass particles of approximately 1/4 of the pipe diameter. Thus 10 mm coal particles jammed a 25 mm flowmeter. In some gas drainage operations this is not an issue but in others it is. Bigger flowmeters do not suffer from these problems.

A simple bi-flap flowmeter can be produced for about $500 and the electronics for a similar cost provided that there is an adequate market so that mass manufacturing techniques may be used. As always in the coal industry the market is small and the cost benefits of mass manufacture are hard to achieve without other industrial applications being found for the devices.

9. CONCLUSIONS

A robust set of communication electronics has been developed to obtain information from a series of sensors along a 2 km long twisted pair power and communications cable. The system is designed to be intrinsically safe and to interface with Sigra's gas flow meters. In addition to the gas flow meters the sensor electronics may obtain data from a variety of transducers. These include potentiometers, hall effect and bridge devices. As such the system may be used to log extensiometers or convergence meters utilizing potentiometers.

The communications system is specifically designed to meet these requirements and does not conform to any of the existing standards such as RS485. The benefits are its particularly low power consumption and simplicity of hook up as it only uses a single twisted pair cable for all power and communications. It has a significant potential use in general monitoring along any mine roadway.

10. ACKNOWLEDGMENTS

ACARP is thanked for its financial support over the long drawn out period of this research project.

The individuals without which this project would not have come to conclusion have been Phillip Eade, Jeff Wood and Greg Eaton all of BHP-Billiton.

John Hanes and Jon Sleeman of ACARP are also thanked for their encouragement.

APPENDICES

APPENDIX 1 - THE ADDRESS DECODER

A receiver contains an address decoder. This consists of a current limiting amplifier feeding the data input of a serial input shift register. The input to the limiting amplifier is the voltage across the two wires. This voltage drop increases and decreases as a transmitter current sink is turned on or off.

The shift register is clocked by an oscillator. The oscillator is started by a pulse from the limiting amplifier output. A counter stops the oscillator after eight bits have been clocked into the shift register. The shift register outputs are compared to the address switches for the device. If a match is found power is applied to the sensor units processor.

APPENDIX 2 - THE SENSOR ADDRESSES AND CALIBRATION

COEFFICIENTS

The sensor contains two addresses, the first of these is defined by the manually adjusted jumpers. This is the decoder address and there must never be more than one sensor with this decoder address on any one system. There may however be more than one system and each of these may have sensors with the same decoder address. This is essential as only 255 such addresses may be used.

The second address is a more involved sensor address which uniquely defines the device. This is a 20 bit number. Each time a sensor is addressed it sends back to the controller its unique number so as to avoid confusion as to which device is being addressed. The sensor also stores in memory the calibration coefficients of the transducers used in the sensor.

These are:

- Temperature - offset and slope constants.

- Hall Effect Sensors and bridge sensors.

 A six segment (seven pairs of constants) calibration table for STP conditions and a four segment (five pairs of constants) calibration table for offset changes with temperature.

APPENDIX 3 - FLOWMETER SIZES

Note: Greater flows may be accommodated for a given size flowmeter but the pressure drops will be greater.

FIGURES

Figure 1- Flowmeter with Hall Effect Sensor

Figure 2- Bi-flap Flowmeter

Figure 3- Block Diagram of System

Figure 4- Block Diagram of Controller