Baker R.C. Flow Measurement Handbook: Industrial Designs, Operating Principles, Performance, and Applications

Подождите немного. Документ загружается.

18.10 CHAPTER CONCLUSIONS 437

PIPE

REFLECTOR

Figure 18.10. Ultrasonic-type insertion probe (after Rawes and Sanderson 1997).

18.9 THERMAL PROBES

Thermal probes are used so frequently as part of a full-bore flowmeter spool piece

that they have been included in Chapter 15.

18.10 CHAPTER CONCLUSIONS

The turbine insertion probe has been an essential workhorse of the industry and

is likely to continue to be, particularly where nonconducting liquids and gases are

involved.

The latest developments in electromagnetic probes are very promising, and fur-

ther designs and developments may well appear.

The novel ultrasonic meter of Rawes and Sanderson (1997) may well set a new

standard for precision in the measurement of bulk flows using insertion methods. It

is likely to be the subject of considerable development, in order to perfect both the

mechanical insertion system and the ultrasonic sensing and averaging procedure.

No doubt a further area of development will be the secondary equipment and its

ability to obtain the best averaging of flow traverse data or of the ultrasonic response.

One of the weakest parts of in situ calibration using insertion probes tends to

be the measurement of internal pipe dimensions. The ultrasonic techniques should

provide a precise means of achieving the cross-sectional area.

The problems of blockage and the effect of wall proximity on the flow around

the probe may lend themselves to computational modeling. However, the ultrasonic

techniques may, again, offer a means to obtain the profile modifications due to the

probe.

CHAPTER

Modern Control Systems

19.1 INTRODUCTION

This chapter does not attempt more than a cursory review of modern control sys-

tems,

as it is outside my main area of expertise. Those knowledgeable in electronics

and control will pass quickly over this chapter! Others who wish to delve more

deeply may, for instance, turn to Sanderson (1988), Brignell and White (1994),

or other experts, or for a good, recent introduction to the control of machinery

Foster (1998) is recommended. I am pleased to acknowledge valuable insights ob-

tained from Cranfield lectures given by Bill Black, David Clitherow, Doug de Sa, and

others.

This book is about flowmeters that depend, in the main, on mechanical or

thermo/fluid mechanical effects. In only one case of mainstream meters, the elec-

tromagnetic meter is the signal electrical throughout. The aim of the book has been

to bring together information on the performance of these meters based on physi-

cal,

experimental, and industrial experience. I have made the assumption that the

electrical signal interpretation, which is based in the meter, is capable of doing the

job set by the designer of interpreting the meter output into a standard electrical

form. This assumption is not always adequate (e.g., where electrical signals from the

same instrument do not appear to agree with each other). Nevertheless, the devel-

opments in design of electrical systems are assumed to be such that the reader will

not benefit from a detailed description or that the circuit will be changing so fast

that any statement will be out of date.

Once the signal is available, we shall need to feed it into the overall control sys-

tem. It is clear that if it does not match the rest of the control system's protocol,

conditioning may be necessary, but this is an expert job, and detailed signal condi-

tioning will not be undertaken lightly. In many

cases,

interface modules that do this

conversion will be available.

From the time when process operators adjusted valves to control flows on the

basis of visual flow readings, electrical control has steadily replaced the operators'

functions. Like the operators, the sensitivity of the electrical systems could result

in undercorrecting or overcorrecting and could cause unstable operation by ampli-

fying the natural fluctuation of the process. The operator then needed to keep an

eye on the automatic control and intervene if necessary. These control systems were

analogue. They used an electrical signal and amplified it, integrated it, and differ-

entiated it in a way analogous to mathematical actions. However, the operator was

always ultimately responsible for setting the values and able to override if necessary.

438

19.1 INTRODUCTION 439

The advent of digital computers allowed the acquisition of much more data and

more detailed assessment of the performance of the plant.

However, as digital computers superseded (largely) analogue computers, so digital

control has superseded analogue control in many cases so that the computer can

respond to signals and adjust the plant in a far more detailed and sophisticated way.

19.1.1 ANALOGUE VERSUS DIGITAL

Signal conditioning is a major subject in its own right, and the reader is referred to

other sources such as Sanderson (1988). Sanderson classifies transducers by

• type of output - voltage, current, or charge;

• type of signal - unidirectional, bidirectional, alternating, or transient;

• information content of signal - by level or by amplitude, frequency, or phase

modulation of a carrier wave;

• relationship between physical variable and electrical output - linear or nonlinear;

and

• output impedance.

Analogue signal conditioning is often to

• convert output to voltage;

• change impedance as required; and

• amplify, filter, demodulate, and linearize.

From classical control theory with its sophisticated mathematics and relative

inflexibility, the modern control systems use the power of the computer to oversee

the process, to question the sensors, to identify problems and rectify them, and to

provide the operator with a vast amount of information and control, but they are

packaged in a way that is easily accessible.

Distributed control systems (DCS), which are analogue, need to be fast with

higher bandwidth and built-in redundancy (e.g., for the petrochemical industry).

In much primarily mechanical plant, the control was originally exercised by

electromechanical relays. These relays would control the electrical power to the

controlling elements in the system (e.g., valves). When these were superseded by

programmable logic controllers (PLC), the logic used in the relays was carried over

as ladder logic.

One most important feature of digital systems is that they are less susceptible to

interference in that they can carry their own checking protocol within the message.

19.1.2 PRESENT AND FUTURE INNOVATIONS

A major step in the move toward modern control was to remove the analogue sys-

tems that were much less flexible, and could not easily be reprogrammed by a com-

puter command, and to introduce the microprocessor. This then allowed a further

sophistication - the individual instruments became intelligent and were able to be

instructed to do simple tasks and, in turn, could provide much more information

to the central computer. The computer could store past information and compare

past performance with current operation, thus providing further checks on plant

performance and safety.

440 MODERN CONTROL SYSTEMS

This led to the idea that the communication system could be a signal highway

(or bus), to which any instrument with the right protocol and any computer could

be connected. The signal highway could be of copper, but equally could be of optical

fiber, electromagnetic waves, or superimposed signals carried by the electricity main

or telephone lines. Transmitters and repeaters could then include satellites, and the

system could be, essentially, unbounded.

At the other end of this communication link is the computer or computers that

control the process and that will be able to work in tandem if necessary, and the plant

operators will work to visual display unit (VDU) screens with the latest windows

presentations and graphical portrayal of each part and aspect of the plant and its

operation.

All this requires a generally accepted common communication language or pro-

tocol to which instruments and systems conform and that will ensure each message

is correctly sent and received, without interference.

Modern telemetry systems provide the means of controlling plant and measuring

its performance from a central station. The three aspects of the process are

• the measurement technology, which has been the prime subject of this book. The

microcomputer has revolutionized this technology with smart and intelligent

instrumentation. One aspect that may need attention is signal conditioning.

The signal from a sensor may not be suitable for a system and may need to be

adapted: from analogue to digital, from digital to analogue, or other change.

• the method of transmission, which, having been by means of copper wires for

much of its history to the present, is now the subject of continual and highly

sophisticated alternative developments.

• the controls, both instrument outputs and process adjustment means. From di-

rect, but remote, control of instruments, through analogue computers, this has

now firmly arrived at digital computer techniques.

19.1.3 INDUSTRIAL IMPLICATIONS

The range of industries is now very wide and includes water and sewage treatment

and distribution, oil and gas production and pipelines, process industries of all sorts,

and manufacturing including the production of flowmeters. The techniques are, of

course, not limited to process flow but can also be used in the electricity genera-

tion and distribution industry, the electrical power transmission industries (e.g., rail

networks), building services, and lighting control.

This means that the typical company supplying such equipment and systems

is increasingly a software-based company, using proprietary equipment and supply-

ing the software for the control station and the outstations, and using whichever

transmission system is most appropriate to the task and the customer's needs.

19.1.4 CHAPTER OUTLINE

This chapter provides an overview of how the signals from these meters are used in

the control of

plant.

In referring to meter or flowmeter, the reader should understand

that broadly the same will be true of other instruments within the system. I shall

start from the assumption that, despite increasing coordination between signal spec-

ifications (protocols), there is still sufficient variation to make it impossible to use

19.2 INSTRUMENT

441

an unconditioned signal from any instru-

ment. The modern control systems will,

therefore, in some form, require the follow-

ing components (Figure 19.1):

a. the instrument;

an interface box that translates electri-

cal signal to and from the instrument

into a signal as required by the remain-

der of the system;

c. a communication language or protocol;

d. a communication medium;

e. an interface that receives the signals

and translates them to the computer

protocol;

f. a computer that receives signals,

checks their validity and appropriate-

ness against limits, models, and past

experience; transmits control informa-

tion to both meters and valves, etc.;

and displays its actions with the maxi-

mum user accessibility.

COMPUTER SYSTEMS

INTERFACE

BOXES

BUS

INTERFACE BOXES

COMMON

PROTOCOL

INSTRUMENTS ON THE

PROCESS PLANT

FLOWMETER

Figure 19.1. Main components required for a modern

control system.

19.2 INSTRUMENT

19.2.1 TYPES OF SIGNAL

The rest of this book is about the details of the flow sensor and how it senses a flow

of fluid and converts this into an electrical signal. The signal will usually be one or

more of the following:

• an analogue signal of 0-20 or 4-20 mA or a voltage or a resistance;

• a frequency signal - this may be of high or low frequency or both;

• a digital signal in which the flow quantity is transmitted in a numerical form,

more likely to be binary than decimal;

• an intelligent communication that allows the sensor to provide more than one

piece of information, and the user, through a computer, to instruct the sensor

(and other devices) in various ways.

Twenty-five years ago we were concerned with the cleanness of the signal, which

came from a meter and provided just one parameter, the flow. The advent of com-

puting power, and particularly the microprocessor, has allowed more information

from the signal to be interrogated and multiple outputs to be handled. Thus interest

is now being shown in the details of the noise in the signal, which previously we

smoothed out and discarded.

442 MODERN CONTROL SYSTEMS

19.2.2 SIGNAL CONTENT

Amadi-Echendu and Hurren (1990) (cf. Amadi-Echendu and Higham 1990, Amadi-

Echendu and Zhu 1992) discussed the use of noise in flowmeter signals to obtain

information about the condition of process plant and condition monitoring in gen-

eral. For instance, the variation of pulse spacing in a turbine meter may be obtained,

and the nature of this fluctuation may be interpreted as due to the signature of

the plant made up of pulsatile flow, etc. Change in the signature may indicate a

change in conditions in the plant. To date, there does not seem to be much theoreti-

cal analysis to help our understanding of the relationship between secondary signal

and the physical fluid behavior. However, some recent work by Krafft et al. (1996)

has started to do just this by developing the theoretical prediction of the effect of

air bubbles in a magnetic flowmeter and comparing this to the signal obtained.

The microprocessor has allowed a further, and major, step. It is now possible to

build into the instrument means for checking the validity of the signal, for adjusting

range, and for selecting the form of

signal.

The meter may, therefore, be capable both

of transmitting and receiving information.

For such meters, the terms smart instrument and intelligent instrument were

coined and associated with various degrees of on-board computational power. The

definition of these terms can be found in Higham and Johnston (1992), but the con-

tinual developments resulting from microprocessor power mean that the distinction

is likely to become less relevant or used. They suggested that smart be applied to sen-

sors that automatically compensate for nonlinearity, temperature effects, etc., and

can be configured before installation to provide a predetermined range and span.

However, they would normally need to be taken out of service to be reconfigured.

The term

intelligent

includes the smart features, but also implies two-way com-

munication with the control system either by a separate communication port or by

a superimposed digital signal on the 4- to 20-mA signal. This greatly increases the

versatility of the sensor, allowing self-diagnostics and other routines and commands

to be sent from the control system to the instrument.

Brignell and White (1994) list the principal subsystems within an intelligent

sensor as

• a primary sensing element,

• excitation control,

• amplification and possibly variable gain,

• analogue filtering,

• data conversion,

• compensation,

• digital information processing, and

• digital communications processing.

A most important attribute of the intelligent sensor is addressability, which is a

necessary part of the bus communication, but it also will put time constraints on

the signal transmission each way. Thus the instrument's microprocessor

• processes signals from one or more sensors and replaces complex analogue

electronics;

19.3 INTERFACE BOX BETWEEN THE INSTRUMENT AND THE SYSTEM 443

• is capable of re-ranging, diagnosing problems, etc.;

• oversees the communication of the sensor with the wider control system; and

• manages display functions if present.

The microprocessor then allows enhanced accuracy by putting in the calibration

curve, which provides

• high turndown ratios leading to increased rangeability - but beware of hidden

errors,

• increased manufacturing flexibility, and

• increased reliability and with a quicker identification of any problems, an im-

proved maintenance philosophy.

There may be some disadvantages such as a change in working practices, the

possibility of changing ranges without considering safety implications, and the sit-

uation of being harder to repair by in-house people and so more likely to lead to

a throw away/replacement philosophy. However, many of these problems have the

solution implicit in the new technology, by using passwords, more sophisticated

algorithms, etc.

19.3 INTERFACE BOX BETWEEN THE INSTRUMENT

AND THE SYSTEM

The system designer and builder is unlikely to be the manufacturer of all the instru-

ments in the system. There was a trend this way about 25 years ago, but the greater

flexibility of the microprocessor has encouraged control system design to be capable

of coping with a variety of instruments from different manufacturers and for various

industries. Thus, although some instruments may incorporate the interface in the

sensor, the simplest arrangement is to have an interface box that ensures a common

signal protocol. This reflects the comparative cheapness of the microprocessor-based

box, in comparison with the cost of expert time to adapt an instrument to a partic-

ular protocol. These devices are sometimes called controllers and may be capable of

handling from 1 to 30 or so sensors with individual loops.

The output of this box (to the highway/bus) will be one of the emerging signal

protocols. This will probably be a digital signal that has a code to describe the nature

of the particular data set that follows the code. The codes will then define particular

instruments (or control devices) to which the signals are directed or from which

they come and, in addition, will define the nature of the data, whether flow rate

etc.,

or range, or some other quality. Assuming that these code and data strings will

be vying for space on the same pair of

wires,

or the same optical fiber, the instrument

will need to set up a "handshake" arrangement with the rest of the system to allow

time for the signal to be transmitted. This in turn will have repercussions for data

transmission rates.

This type of outstation may also provide local control and may have main and

standby units.

444 MODERN CONTROL SYSTEMS

19.4 COMMUNICATION PROTOCOL

19.4.1 BUS CONFIGURATION

In the past, connection of sensors to the central control has been in the form of

a star where each instrument communicates via a, usually, wire pair to the center

(Brignell and White 1994). There is clearly much cabling required, and it becomes

most dense at the center. The modern alternative, the bus, allows all devices to share

a common pair of wires. Disadvantages are that

• devices need to be addressable and so to have a level of intelligence,

• breakage in the bus can cause a loss of communication, and

• pressure on the bus may limit time availability for signal transmission.

To overcome the particular danger of bus breakage, a complete loop bus can be

used, or even two bus loops separated physically to avoid the common destruction

of both.

The system components will use one of the standard protocols, for instance:

• 4 to 20 mA signal for analogue instruments;

• digital instruments with digital communication;

• flow computer output with serial interface that provides an encoded binary trans-

mission; and

• a modern bus with high level, computationally controlled protocols (e.g., HART,

Profibus, and Fieldbus), which interface to intelligent instruments.

It should be remembered that digital information in computers is carried as a

series of Is or 0s and so, for instance, the logic 1 could be portrayed as a burst

of higher frequency, whereas the logic 0 would be portrayed as a burst of lower

frequency. The message would consist of a series of Is and 0s, with indications of

start and finish of words, and any instrument attached to the cables would respond

when its call sign was sensed on the bus/highway.

Brignell and White (1994) discussed the ISO's (see ISO 7498-1:1995) seven-

layered communications protocol for an Open Systems Interconnect (OSI) strategy.

Physical layer Relates to connections such as physical

connectors, and electrical procedures for

handshaking, etc.

Data link layer Ensures reliable and error-free

transmission.

Network layer Appears to ensure standard signals and

appropriate communication paths regardless

of where the signal originates.

Transport layer Provides a reliable data transfer exchange

service between the instruments, etc., and

the user.

Session layer Manages the interaction between different

applications on different systems

connected to the network.

19.4 COMMUNICATION PROTOCOL 445

Presentation layer Ensures a common data format.

Applications layer Provides services that can be accessed

and used by the computer systems.

These cover, broadly, the levels (a) to (f) in Section 19.1.4.

19.4.2 BUS PROTOCOLS

Manufacturing Applications Protocol (MAP) is based on the seven-layer ISO/OSI

model. General Motors was largely responsible for its development (Brignell and

White 1994). Statistical tasks for optimizing production, order processing, planning,

and different logistical services may be available from such protocols (Lindner 1990).

However, MAP is not widely used as a general-purpose instrumentation standard due

to the severe requirements for the hardware/software processing resulting from the

full seven-layer protocol.

Wood (1994) described HART and Profibus as examples of proven specifications

that provided background and influence for the modern fieldbus proposals.

Highway Addressable Remote Transducer (HART) digital communications pro-

tocol developed by Rosemount (Brignell and White 1994) uses digital techniques to

allow an intelligent instrument to adjust range and span, to undertake flow calcu-

lations, to apply self diagnostics, and to communicate (Howarth 1994). HART uses

point-to-point or all-digital modes. In the former, it superimposes digital data on a

4 to 20 mA signal by two defined frequencies of ±0.5 mA amplitude: 1,200 Hz for

binary 1 and 2,200 for binary 0. Message rates are typically three per second. It uses

a master/slave protocol (Brignell and White 1994) in which the master is the only

device that may initiate a message transaction. Up to two masters can communicate

with connected field instruments, typically a hand-held communicator and a con-

trol system or PC-based workstation. At the applications layer, HART commands are

of three types: universal commands to read an identifier, variable, or current; com-

mon practice commands (e.g. calibrate or perform self

test);

and transmitter-specific

commands (e.g., start/stop totalizer, read density calibration factor).

Profibus was set up by a group including AEG, ASEA Brown Boveri, Honeywell,

Klochner-Moeller, Lands & Gyr, Robert Bosch and Siemens, with the objective of

creating a digital fieldbus standard consistent with the ISO/OSI model in ISO 7498

(Squirrell 1994).

Interoperable Systems Project (ISP) was formed in 1992 (Allen 1994) with the

objective of achieving a single, open, interoperable, international fieldbus.

World FIP (Desjardins 1994) was to promote the acceleration of a viable interna-

tional fieldbus standard of manufacturing and process control and resulted from a

memorandum of intent in 1993.

The ISP and World FIP (North America) combined in 1996 to form Fieldbus

Foundation, which produced in 1998 an interoperable replacement for 4 to 20 mA

control, that is at least as reliable and replaceable. Members represent instrument

suppliers and end users.

Fieldbus is a communication system specially designed for networking of trans-

ducers (Brignell and White 1994). The protocol model has physical layer, data link

layer, applications layer, and also a user layer.

446 MODERN CONTROL SYSTEMS

In conclusion

should

noted that this brief introduction

fieldbuses

was

believed

to be

correct

at the

time

writing,

but it

should also

remembered that

there

are

other fieldbuses available

addition

those mentioned

this section,

and

the

reader should obtain expert advice

on the one

most appropriate

to a

partic-

ular application.

19.5 COMMUNICATION MEDIUM

19.5.1

EXISTING METHODS

TRANSMISSION

Pneumatic transmission

has

been used widely

in the

past

and had the

great virtue

of being intrinsically safe.

with analogue electrical signals, standard signal ranges

ensure that instruments are compatible

and

that pressure supply failure will

not

pro-

duce

undetected error. Thus 20-100 kPa

and

20-180 kPa are two such (Sanderson

1988).

It is

likely that with optical transmission offering

alternative and,

most

cases,

intrinsically safe communication system,

and

with

the

inherent compatibil-

ity

optical with digital electrical systems, pneumatic systems will become less

common.

Analogue electrical transmission depends

a small current, voltage or resistance

to provide

the

signal from

sensor. Much instrumentation still

service uses these

signals,

and so

analogue communication

likely

to be

around

for

some time. Again

likely

decrease,

and,

indeed, sensors that exploit digital

and

other modern

technologies

are

likely

to be

designed. Many analogue sensors included

frequency

output, which allowed pulses

to be

counted

at the end of the

communication line

and

total flow

other parameter

to be

obtained.

is now

possible

send complete information

parameter values

digital

signals, which

do not

need

to be

converted from

milliampere value

to a

flow rate

but which provide

the

flow rate

in the

required units.

addition,

the

sensor

can

be addressed

and

will respond

questions

commands.

The

communication

has

thus become two-way.

19.5.2 PRESENT

AND

FUTURE TRENDS

The means

which

the

components

in a

modern control system

are

connected

are

now

many

and

various

and

exhibit some

of the

ingenuity that stems from

the

great power

modern signal processing. Thus communication

for a

conventional

plant

may

well

be via

copper wires

or,

where there

are

dangers

due to

explosive

environments,

via

optical cables.

The

transmission system

may be by the

public

switch telephone network (PSTN). A frequency signal

essentially

the

signal used

in telephone communications and, therefore,

most

the world's communication

all

forms

transmission

and in

repeater stations such

satellites.

For oil and

gas,

transmission

may be by

microwave, which requires line

sight between

the

control station

and

outstation,

or by

radio waves bounced

off the

troposphere

linked

satellite.

utility

may

develop means

for

sending

the

signals down

the

power lines that link every house, whereas

international

oil or

gas company may

be able

afford radio links

via

their

own

satellite

rented satellite space.