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

Подождите немного. Документ загружается.

19.5 COMMUNICATION MEDIUM 447

Combinations of these and other methods will be selected depending on the

most appropriate route for the particular application. For added security, dual meth-

ods may be used providing a fall-back in case one method fails. The options are likely

to increase.

The bus/highway will increasingly link computers with control and communica-

tion capability and smaller modems with hand-held devices for field checking. The

development of such data highways is likely, eventually, to lead to the elimination

of all but digital information. Standardization will address the frequency levels of

the bus, which will affect the speed of response of the instruments.

The sensor will, therefore, be capable of linking with computer control systems

anywhere in the world or beyond and the wire connections may become an antenna.

Even the problems of power supply are overcome with local power generation from

solar, wind, or nuclear sources.

19.5.3 OPTIONS

Imrie (1994) commented that the type of measurement being made will determine

to a large extent the type of communications required:

• Is the variable to be used in a control or a time-critical application?

• What is the likely rate of change of the variable?

• Is continuous trend monitoring required, or is it event-driven?

• Does the sensor lend itself to power saving?

The location will determine security, protection, safety, interference, availability

of services (e.g., power and signal networks), and the access requirements of owner-

ship of land.

The consequence of data loss, the rate of data transfer, the response time, one-

way or two-way transfer, and the size of data package all will affect choice.

The nature of the application will be important: control or monitoring, regula-

tory and insurance, safety, effect on product or services.

Imrie then went over the options: cable direct connected, service radio, satellite

communications, mains borne, meteor burst, licensed radio, and delicensed low

power radio. In addition, careful consideration of networks and protocols and of

power management is necessary. Imrie suggested the following applications:

For Use

Real-time measurement Cable or radio

and/or control

Time critical Cable or radio

Event-driven with a Cable or radio with

fast response appropriate protocol

Event-driven with a A link establishment

slow response network over telephone or

service radio networks, or

low use power radio

448 MODERN CONTROL SYSTEMS

Long-term trend measurement Link establishment network

over telephone or service

radio networks with overnight

historical file transfer

Multinode system Suitable network, not multiple

point-to-point links

19.6 INTERFACE BETWEEN COMMUNICATION MEDIUM

AND THE COMPUTER

For a particular company designing a control system, this interface will be essentially

a modem that converts the electrical signal transmitted to the size and form needed

by the computer. The transmitted signal will be specified by the protocol or by the

communication network management, and the other side of the interface should

be under the control of the system designer and meet the requirements of the

computer interface.

19.7 THE COMPUTER

By comparing the operation of the process with models programmed into the com-

puter, the system could set the control points, removing operator error and reducing

operator time.

There is wide use of modeling and simulation techniques for areas as diverse as

network analysis for water distribution systems, chemical mixing, and pump control

using software suites (Fowles 1994).

The control station, which will represent the latest in user-friendly presentation,

will probably include a computer model of the network, based on predictions such

as those of Miller (1990). Thus, the computer will enable the user to identify on a

map where the control is taking place, will provide the user with a virtual control

panel, will allow the user to obtain trends and to compare them with expectation,

and will provide alarm controls, etc.

There are over 100 suppliers of SCAD A/telemetry packages (Fowles 1994). Super-

visory control and data acquisition (SCADA) systems send data from the field to the

model, and the model assesses the data to distinguish good from bad. Supervisory

control will include emergency shutdown, with a block valve in an automatic con-

trol loop. Thus in the water industry, there may be local automatic control shutdown

of the pumping system, which operates against, say, the depth of water in the tanks.

In the oil and gas industry, this may not be the case.

Microprocessor-based protection systems (MBPS) are usually concerned with an

undesired end event (Churchley 1994/5).

19.8 CONTROL ROOM AND WORK STATION

Little needs to be added to comments already made. It will be obvious that this

area will increasingly represent the most modern office with well-laid-out computer

19.11 FUTURE IMPLICATIONS OF INFORMATION TECHNOLOGY 449

equipment readily available to the operator, encouraging able people with a high

level of technical training and interest to operate the plant with a suitable level of

responsibility.

19.9 HAND-HELD INTERROGATION DEVICE

The development of hand-held devices is particularly important in the following

two areas:

a. As a device for the process plant engineer enabling him or her to interrogate the

system while at various points on the site. Lindner (1990) suggested that these

devices will usually be battery powered and will be used for identification, con-

figuration, and diagnostics and used the graphical title of digital screwdrivers!

These instruments will need to be capable of operating in adverse environments.

Providing the utility industry with instruments that can interrogate the utility

meter via a remote reading pad using inductive or other pickup, so that, without

entering the property, the meter reader can obtain meter identification and

logged information. In due course, these may give way to signals transmitted,

say, by telephone or radio links.

19.10 AN INDUSTRIAL APPLICATION

Boettcher and Hickling (1993) described the systems for Sizewell B power station

in the United Kingdom. The man-machine interface was designed to ensure that all

information was available to the operators with high reliability. The user interacts via

graphical screen displays of windows, icons, menus, and pointers, a so-called WIMP

environment. This results in an extensive interactive capability of the computer

system. Alarm systems are available in various formats. Safety and reliability are

fundamental to such a system.

distributed computer architecture was used to allow

flexibility and extendability.

Communication was over local networks. The data processing and control

systems were divided into two levels. Level 1 comprised the set of programable logic

controller type devices and interacted directly between devices and the control

room panels, reading sensor and status of devices and issuing control signals. These

were high integrity control systems (HICS). Information was transmitted to level 2

distributed control system (DCS) for display on the VDUs. Separation, duplication,

and redundancy were built into HICS (cf. Cluley 1994/5, who discussed reliability,

the use of redundancy, environmental effects, self-calibration, display devices,

software reliability, component selection, environmental factors, and measures of

reliability).

19.11 FUTURE IMPLICATIONS OF INFORMATION TECHNOLOGY

From a mechanical engineer's viewpoint, there appear to be a number of fairly obvi-

ous ways in which the power of information technology will continue to influence

450 MODERN CONTROL SYSTEMS

the instrumentation and control field. These are likely to be driven by the rapidly

advancing computer technologies.

Control Station

Leaving aside the possibilities that humans may communicate with the computer

by means other than hands, feet, and voice, the visual graphics, the virtual reality

developments, and the windows environment of modern computers will demand

of the process engineering designer increasingly detailed simulations of hardware,

software, physical behavior, and safety scenarios with chaos theory implications. In

many cases, the expert system will provide the operator with an instant assessment

of the most likely causes of failure or of unusual operations.

Behind such information will be increasingly detailed computer graphics but,

more importantly, increasingly detailed understanding of the complex nature of

flow phenomena. The computer experts, with engineers and scientists, will attempt

to model various theories of catastrophic failure to enable the computer system to

recognize possible symptoms at an early stage.

Communications

The move to digital has released most constraints from the transfer of information

and opened up the vast array of current communications used in telephone, space,

etc.

This is likely to yield new and unforeseen additional possibilities.

However, it is possible that, in some cases, the work on neural communication,

and the methods used there, may introduce faster links in the communications chain

or new statistical insights into information transfer.

Sensors

Most sensors in use today stem from the mechanical engineer's need to measure

quantities in the most efficient ways prior to information technology. If one starts

from the power of the microprocessor, and asks how one would design a sensor, the

answer may be far more complicated computationally, but extremely simple and

basic from a mechanical viewpoint. Already we are seeing a move to microsensors

that can be built into chips. These depend on thermal and vibrational techniques.

Other chemical or tactile effects may be exploited in the near future.

An information technology specialist may offer many other possibilities, but the

instrument engineer, be he or she inventor, designer, manufacturer, entrepreneur,

or chief executive, should not wait for such suggestions but rather be prepared to

think the unthinkable. In the short term, this may appear to be a lessening of the

importance of mechanical engineering in instrumentation, but in the longer term it

is likely to lead to an enhancement of the instrument engineer's ability to measure

the real world.

CHAPTER

Some Reflections on Flowmeter

Manufacture, Production, and Markets

20.1 INTRODUCTION

I do not claim an expertise in manufacture, but I want to make three points briefly

in this chapter and then to enlarge on the last of them.

Market information is usually inaccessible, and there may be a case for research

centers to work together to provide a better source of data for industry. It is clear

that there is not always an entirely satisfactory information flow, even within com-

panies, between the marketing department and the technical and manufacturing

operations.

My first point is, therefore, to suggest the need for better market information

both within and outside companies.

My second point is to indicate the need for the flow metering industry, on the

one hand, and the science base as it relates to instrumentation, on the other hand,

to create an effective means for technology transfer, which in the process will raise

the profile of the sector and ensure that government is aware of its importance.

Past experience of encouraging collaboration suggests that industry and the science

base do not always appreciate the value of working together. Collaboration between

industry and the science base should be mutually beneficial in providing an antenna

for new information on developing technologies.

The third point is that the production of instrumentation is a special case in

which the effect of the production process on the final accuracy of the instrument

may, in some cases, be predicted and used to specify the production requirements.

Although a special case, it may be important for other products; the ability to mea-

sure dimensional precision and variation at each stage of the production may lead

to higher quality in other products.

20.2 INSTRUMENTATION MARKETS

The flowmeter has been described as the cash register of the process industry. The fol-

lowing figures give some indication of the turnover. Information on meter markets

tends to be so commercially valuable that it is treated as highly confidential. Individ-

ual firms are unwilling or, for commercial reasons, unable to reveal their own knowl-

edge.

This section, therefore, is gleaned from information in the public domain.

Halsey (1986) obtained a 20% response to a questionnaire sent to 330 companies

thought to be using flowmeters. He analyzed the resulting data in various ways. He

451

452

SOME REFLECTIONS ON FLOWMETER MANUFACTURE, PRODUCTION, AND MARKETS

obtained the percentage of meters used by industry type.

Fluid Industry Type

Liquids Water

Hydrocarbons

Other chemicals

Food

Other

Total

Gases Air

Steam

Hydrocarbons

Other chemicals

Other

Total

Mixtures Total

Share of Flowmeters (%)

Negligible

ihown here, the meters used were predominantly orifice pla

Orifice plates

Other differential pressure

Variable area

Positive displacement

Turbine

Vortex

Electromagnetic

Ultrasonic

Mass

56%

19%

Negligible

Of these, 66% were owned by large users (defined as having over 600 meters),

and only 4% were with small users (with 60 meters or fewer). Of the meters, 83%

were used for continuous flow, whereas 8% were used for batch processing; 5%,

for custody transfer; and the remaining 4%, for research and development (R&D)

or as flow alarms. He concluded that positive displacement meters were in decline,

whereas electromagnetic meters were increasing market share. Vortex and mass were

increasing, and smaller users were more likely to have a greater proportion of newer

types.

He made no estimate of the total market but noted that about 70 companies

used over 17,000 flowmeters.

No doubt the situation has changed since 1986. More recently, Lawton Smith

(1994) has given some overall figures. These are approximately as follows:

Flow measurement

industry worldwide

Value of oil and gas

monitored by flowmeters

• in the United Kingdom

Total of all industrial sectors that depend

on flow metering

• in the United Kingdom

• worldwide

$700-800 million p.a.

>$30 billion p.a.

$250 billion p.a.

$10,000 billion p.a.

20.3 MAKING USE OF THE SCIENCE BASE 453

These figures should probably be increased by about 5% p.a.

;

suggesting a world-

wide flow measurement industry of US $1 billion or more by the year 2000. Some

may suggest that this is a low estimate, but it will depend on the secondary instru-

mentation, etc., which is included in any estimate.

She used gas meters as an example of utility markets in the three countries that

she studied:

Market Number Produced p.a.

United Kingdom 2 million

France 400,000

Belgium 60,000-80,000

Lawton Smith estimated the following comparative figures for numbers of man-

ufacturers and suppliers:

United Kingdom

France

Belgium

200

A DTI report (1996) suggested that, of the total instrumentation and control

industry in the United Kingdom, about 40% is in process monitoring and control.

Of the six major producing countries, breakdown of the total instrumentation and

control industry production between them was as follows:

United States

Japan

Germany

United Kingdom

Italy

France

36%

24%

8.5%

2.5%

These figures allow an approximate estimate to be made of the likely market and

the competition worldwide. Within the flowmeter market, at the time of writing,

Rosemount was probably the dominant producer, with a range of competitors who,

typically, were international companies with operations around the world, often

each national branch having a particular flowmeter expertise and flowmeter type or

application.

20.3 MAKING USE OF THE SCIENCE BASE

Lawton Smith (1994) identified various issues relating to technology transfer in the

United Kingdom and other countries. The first was the gap between the products

of the research base and the needs of the prospective manufacturer. There is in-

creasing recognition of the need for funding the developments in this gap, and

the public purse is often unwilling to do so. This gap could also be reduced by fa-

cilitating exchanges of people between industry and the science base. Some may

contend that industrial research in the science base may inhibit the generation of

new ideas. However, in its place, it is an essential part of ensuring that engineering

454 SOME REFLECTIONS ON FLOWMETER MANUFACTURE, PRODUCTION, AND MARKETS

and applied science research is focused on the real problems of industry, and this

cross-fertilization will probably generate new ideas.

20.4 IMPLICATIONS FOR INSTRUMENT MANUFACTURE

claim an interest but not an expertise in the manufacture and production of flowme-

ters and other instrumentation. The comments that follow are derived mainly from

observations I have made during visits to companies and from reading and thinking

around the subject. Consequently, they are perhaps rather idiosyncratic. However, I

have benefited from the books by Hayes et al. (1988) and Slack et al. (1995).

It is widely recognized that the design and implementation of a new product

must take account of, and be done in close collaboration with, the production,

manufacturing, and marketing functions. Hayes et al. (1988) made this point when

they suggested that it is preferable for the upstream product development and the

downstream process development to take place linked together by continuous in-

formation exchange. Each must be confident enough and prepared to risk changes

as they develop a product together. This requires skills, appropriate attitudes, and

mutual trust.

Muhlemann et al. (1992) stated, in the context of TQM (Total Quality Manage-

ment),

that "For an organization to be truly effective, every single part of it, each

department, each activity, and each person and each level, must work properly to-

gether, because every person and every activity affects and in turn is affected by

others."

Slack et al. (1995) denned "Quality [as] consistent conformance to customers'

expectations." This requires, for the instrument manufacturer, that the design pro-

cess (possibly based on a computer model of the meter) and the production process

lead to the precision, accuracy, and price expected by the user.

20.5 THE SPECIAL FEATURES

OF THE INSTRUMENTATION INDUSTRY

Modern instrumentation frequently is denned by a precise analytical expression or

algorithm. The output signal should, therefore, be related to the measured parameter

with a high level of confidence. This confidence comes not only from the design but

also from the quality of manufacture. It is this combination that creates a high

quality instrument, in other words, an accurate instrument with a small value of

signal uncertainty.

is,

therefore, reasonable to assume that we can deduce from the manufacturing

process the effect it has on the parameters that control precision and should be able

to identify a method of production of the instrument, which will ensure that the

finished product is of specified quality. We should be able to use an error equation

for changes in each parameter.

This could be used forward or backward. We might establish a value of pre-

cision in measurement of time difference necessary to achieve an accuracy for

the flowmeter and design the time measurement system accordingly. Or we may

20.6 MANUFACTURING CONSIDERATIONS 455

establish the likely value of time error for a given design and construction and apply

this to achieve the uncertainty in flow rate.

20.6 MANUFACTURING CONSIDERATIONS

We review briefly various aspects of flowmeter manufacture relating to these points

or emphasised in the literature in relation to instrumentation.

It is essential to think through the whole product development in detail before

you start so that problems and changes will be kept to a minimum, (see Finkelstein

and Finkelstein 1994 on requirements engineering). Preliminary work must look at

the scope for innovation as well as the constraints: commercial, contractual, proce-

dural, or legal.

There must be a clear understanding of product development, together with valid

reasons for undertaking it, in order to define:

• the specification of the system;

• the management of the development;

• the documentation of the development process;

• rigorous project cost management, scheduling, and decisions relating to changes;

and

• parameters by which progress will be measured against predicted time and costs.

Those with a stake in the process need to be identified and involved from the

start. The successful development will require excellent communications between

the team and those with whom they need to liaise.

This can be achieved by setting up multidisciplinary teams in order to remove

the need to transfer information from one function to another (cf. Constable 1994

on concurrent engineering). Thus the sequential system of concept to R&D to

development to production planning, etc., is broken down with great savings in

terms of time to market. The team given the task of developing a product should be

full-time. This may cause staffing problems when first introduced but, in the longer

term, should be less intensive in staff time. The team should have a leader to whom

the team answers, who also ensures a clear plan and design

brief.

The result of this

is likely to be much increased responsibility pushed down the organization, with

beneficial effects on staff motivation.

20.6.1 PRODUCTION LINE OR CELL?

The observations in this chapter are based on flowmeter production methods. It is

apparent that much flowmeter production is custom-made, responding to the par-

ticular requirements of each customer. Clearly a few flowmeter ranges are produced

in large quantities on a production line (e.g., domestic gas meters).

The conclusion of this is that, for many production operations, a cell system will

offer a useful way to respond to the custom-made nature of the business but will also

provide a more satisfactory working arrangement and a delegation of responsibility

for quality, method, and delivery to the leader and members of the cell.

456 SOME REFLECTIONS ON FLOWMETER MANUFACTURE, PRODUCTION, AND MARKETS

The nature of much instrument production is based on designs of advanced the-

oretical and experimental refinement. As a result, the considerations of production

layout, materials purchase, storage and use, and a fully computerized production

control may take second place to design. However, with modern computer control,

the opportunity should be taken to optimize the production system and to install

a tracking system that will document the details of the production and calibration

on every instrument. The assessment of the needs of a particular manufacturer is

a task that some universities (e.g., Cambridge) find it useful to give to their able

final year students, to the mutual benefit of the student, the university, and the

company.

20.6.2 MEASURES OF PRODUCTION

Those involved in the design of flowmeters will appreciate the need to measure in

other areas, and the need to do so in the production system should come as no

surprise. In many cases, this requires an intelligent appraisal of the system and a

clear understanding of what we are seeking to achieve. However, some useful tools

such as benchmarking the company's performance against companies in the same

business and measuring productivity and other calculable aspects of the production

process are available.

brief discussion of these tools follows.

a. Productivity. Hayes et al. (1988) described an approach that measures improve-

ments in conversion of resources into products. It uses a measure - the single

factor productivity - for the ratio of product output to resource input. The sum of

these factors, weighted according to the amount of resource used in the product,

provided total factor productivity, and its improvement could then be measured

with time.

Benchmarking. Hayes et al. (1988) encouraged benchmarking by identifying

competitors and noncompetitors who have a superior performance/product and

by using trade journals, annual reports, visits, etc., to focus on the reasons for

performance differences.

c. Control charts. Control charts will plot key measures of production with time.

They may measure process quality; defective items; process production times,

etc.;

product precision; and mean time to failure of products.

In the next section, we consider one such measure, particularly appropriate to

instrumentation, where a key factor is final accuracy and where the link between

production process and accuracy can sometimes be analyzed.

20.7 THE EFFECT OF INSTRUMENT ACCURACY

ON PRODUCTION PROCESS

Variation in manufacture, in many cases, is due to wear or operator variation. Al-

though this will also be true for instrumentation, I would like to suggest an addi-

tional reason: small variations in production precision and materials. I shall first

indicate this by descriptive examples and then offer some simple mathematical

examples.