Energy Management System and DDC

7/31/2019 Energy Management System and DDC

1/244


2/244

iii

Energy

Management Systemsand

Direct Digital Control

Richard A. Panke, CEM

MARCEL DEKKER, INC.

New York and Basel

THE FAIRMONT PRESS, INC.

Lilburn, Georgia


3/244

iv

Library of Congress Cataloging-in-Publication Data

Panke, Richard A.Energy management systems and direct digital control/Richard A.

Panke.p. cm.

Includes bibliographical references and index.ISBN 0-88173-395-4 (electronic)

1. Buildings--Energy conservation. 2. Buildings--Electric equip-ment. 3. Digital control systems. I. Title.

TJ163.5.B84 P34 2001658.2--dc2l

2001023849

Energy management systems and direct digital control/Richard A. Panke.2002 by The Fairmont Press. All rights reserved. No part of thispublication may be reproduced or transmitted in any form or by anymeans, electronic or mechanical, including photocopy, recording, orany information storage and retrieval system, without permission inwriting from the publisher.

Fairmont Press, Inc.700 Indian Trail, Lilburn, GA 30047tel: 770-925-9388; fax: 770-381-9865http://www.fairmontpress.com

Distributed by Marcel Dekker, Inc.

270 Madison Avenue, New York, NY 10016

tel: 212-696-9000; fax: 212-685-4540

http://www.dekker.com

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

0-88173-395-4 (The Fairmont Press, Inc.)0-8247-0920-9 (Marcel Dekker, Inc.)

While every effort is made to provide dependable information, the publisher, authors,and editors cannot be held responsible for any errors or omissions.


4/244

v

Contents

Chapter Page

1 Introduction.................................................................................. 1

2 EMS Overview ............................................................................ 9

3 Hardware/System Components ............................................ 15

4 System Architecture .................................................................. 27

5 Direct Digital Control (DDC) ................................................. 35

6 Networking ................................................................................ 47

7 Software/Application Programs ............................................ 65

8 Communication Protocol ......................................................... 79

9 Operator/Machine Interface ................................................... 99

10 Savings/Cost Estimating ........................................................111

11 Sequence of Events ..................................................................119

12 Selection/Expansion ............................................................... 125

13 Installation/Commissioning ................................................. 129

14 Training/Operation/Maintenance ....................................... 13715 Fire Alarm/Security ............................................................... 143

16 Design/Drawings/Specifications ......................................... 147

17 Intelligent Buildings ............................................................... 175

Appendix A Glossary of Terms............................................................ 179

Appendix B Controls Symbols ............................................................. 191

Appendix C EMS Manufacturers ......................................................... 197

Appendix D List of References............................................................. 199

Appendix E Metric Conversion Guide ............................................... 200

Appendix F Sample EMS Problem ..................................................... 203

Appendix G EMS Articles ..................................................................... 209

Appendix H Remember! ........................................................................ 228

Index .................................................................................................... 233


5/244

INTRODUCTION 1

1

Chapter 1

Introduction

EMS DEFINED

MS defined: A system which employs microprocessors, buildingfield panels, communication cables between field panels, controlequipment, and software application programs configured into

a network with control functions at multiple locations and a point ofoperator supervision and control (see Figure 1-1).

Central EMSs are of various levels of sophistication depending onthe size of the building and desired operational function. The simplest

system allows an operator to check the operational status of the heating,ventilating and air conditioning (HVAC), fire and security systems, andcontrol various equipment remotely from a central console. The nextlevel system (direct digital controlDDC) has a digital computer toperform most of the work normally done by the operator, plus otheroptimization and control functions.

Previously, EMS terminology included statements such as com-munication between the field panels and the Central Processing Unit(CPU). Current EMSs no longer require a CPU, nor do they rely on aCPU for system operation!

CONVENTIONAL PNEUMATIC CONTROLS

Conventional pneumatic controls have been the traditional form ofcontrol used in most commercial and institutional facilities for environ-

E


6/244

2 ENERGY MANAGEMENT SYSTEMS AND DIRECT DIGITAL CONTROL

Figure 1-1. Pneumatic to Automation Via DDC


7/244

INTRODUCTION 3

mental control. The control function is performed by a pneumatic con-troller which receives its inputs from pneumatic sensors and sends con-trol signals to pneumatic actuators (see Figure 1-2).

In the 1970s, a supervisory system was often interfaced to thepneumatic control system to allow remote control (remote set point orcontrol point adjustment) of pneumatic receiver controllers and central-ized monitoring through electronic sensors. The remote electronic panelscommunicate to a central computer; however, if the communicationsline was broken, or if the computer failed, the entire system was OFF(see Figure 1-3).

DIRECT DIGITAL CONTROL (DDC)

Direct Digital Control (DDC), although used for years in the pro-cess industry, entered the HVAC industry in the late 1980s. DDC utilizesa programmable microprocessor as the primary controller. The HVACsystem variable (temperature) is measured by electronic sensors. Thecontrol functions are performed by a microprocessor which transmits anelectronic or transduced pneumatic control signal directly to the con-trolled device (damper or valve actuator).

DDC is a form of closed-loop control. The term Direct means themicroprocessor is directly in the control loop and the term Digital

means control is accomplished by the digital electronics of the micropro-cessor. As opposed to electronic controls, which are much like pneu-matic controls where each controller handles one control loop in a fixedmanner, DDC can control numerous control loops and be repro-grammed for different control functions without hardware changes (seeFigure 1-4).

Distributed DDC consists of several DDC units located throughouta building complex. Although, each DDC can operate independently,they are all connected to a central operator station for centralized controland monitoring. A measure of a true distributed DDC system is whetherthe remote DDC units continue to perform full control and energymanagement, without the central operator station (see Figure 1-5). DDCwill be covered in greater detail in Chapter 5.

Refer to Table 1-1 for a comparison of pneumatics and DDC.


8/244


Figure 1-2. Pneumatic Control

Figure 1-3. Supervisory Automation System


9/244

INTRODUCTION 5

Figure 1-4. Direct Digital Control


10/244


Table1-1.Comparison

ofPneumaticsandDirectDigita

lControl

COMPARISON

CONVENTIONAL

BESTCONTROL

CATEGOR

Y

PNEUMATICCONTRO

LS

DIRECTDIGITALCONTROL

SYSTEM

Performan

ce

Proportionalcontrolonly.

FullPIDcontrolandmore.

DDC

Singleloopcontrollers.

Multi-loopcontroller.

Complexcontrolisdifficult

Easytodefinecomplexsequences.

orcostly.

Closercontrol.

Adequatecontrol.

InitialCost

Costriseswithnumber

of

OncecostofDDCcontrolleris

Comparable

controlloops.

absorbed,costriseswithnumberof

Complexcontrolisvery

sensorsandactuators.

expensive.

Capableofmostcomplex

control.

Reliability

Provenreliabilityovermany

Provenreliabilityinproce

ss

DDC

years,however,control

system

industryandmanycomm

ercial

mustbewellmaintainedand

HVACapplications.

recalibratedregularly.

EachDDCcontrollercan

standalone.

Reliesonairsupply.

Maintaina

bility

Relativelyeasytomaintain.

Automaticas-builts.

DDC

Requireregularrecalibration

Built-indiagnostics.

duetodrift.

Fewercomponents.

Nodrift.

Servicebyboardreplacem

ent.


11/244

INTRODUCTION 7

Table1-1.ComparisonofPneumaticsandDirectDigitalContr

ol(C

on

tin

ue

d)

COMPARISON

CONVENTIONAL

BESTCONTROL

CATEGOR

Y

PNEUMATICCONTRO

LS

DIRECTDIGITALCONTROL

SYSTEM

Flexibility

Changesoradditionsre

quire

Programmablecontroller.

DDC

newordifferentcontrollers

Newcontrolstrategiesdefinedatcentral.

re-pipingandoftenwiring,and

Newcontroleasilyadded

.

thenrecalibration.

EaseofUse

Alloperatorinteraction

at

FullEnglishlanguagerep

orts.

DDC

localcontrolpanels.

ColorGraphicDisplays

Canreadtemperaturesand

AutomaticRecordsofall

control

changeset-point.

strategies.

LifeCycle

Cost

Requiresregularrecalib

ration.

Easytomaintain.

DDC

Modificationandexpan

sion

Easytomodify.

requireadditionalcontrollers.

Easytoexpand.

CosttoAddEnergy

Eachnewfunctionusua

lly

Newfunctionsareeasilydefined

DDC

Management

requiresadditionalequipment

byoperator.

andlabor.


12/244



13/244

EMS OVERVIEW 9

9

Chapter 2

EMS Overview

nergy conservation through management has been, and re-mains, one of the most viable energy resources available to allsectors of the energy consuming building community. By mini-

mizing energy consumption and still maintaining the posture requiredfor our business activities, we can save money and therefore survive ourrespective market areas.

An EMS is one energy conservation alternative that can provide ameans to control, reduce and perhaps eliminate energy waste.

HISTORY

There are a multitude of EMSs on the market today ranging fromresidential EMS to large facility management systems. Several types ofEMS are available. Most major control firms and other companies in thisfield have introduced families of building automation systems intendedfor a wide range of building sizes. These systems can be classified asfollows.

Class I SystemsThese consist of small monitoring and control systems that can be

used in buildings with floor areas up to about 100,000 ft2. The basiccomponent is a microprocessor preprogrammed to start/stop different

E


14/244


HVAC system components according to a preselected schedule. Thesystems can be designed to perform other operations such as monitoringfire alarms and smoke detectors, security checks, and load cycling.

Class II SystemsThese systems are similar to those in Class I except that they can

serve larger buildings and some building complexes. The available soft-ware packages provide functions such as: executive and operating in-structions, scheduled start/stop operations, load rotation and shedding,control points resetting, optimization of start time, enthalpy optimiza-tion, and fire alarm and life-safety system monitoring.

These systems can usually monitor about 2000 addressable points.When these systems are used for a group of buildings or building com-plexes, the central control facility is connected to remote data gatheringpanels by means of one or more types of data communication links.Because more than one data gathering panel is served by a central facil-ity, each panel is allotted an equal amount of time in direct communica-tion with the central facility.

Class III SystemsThese are referred to as direct digital control (DDC) systems, and

are the most sophisticated type of EMS. DDC systems are used for build-ing complexes such as medical institutes and university campuses. In

addition to the basic functions described earlier, it is possible to includethe following programs: reset of supply air system; optimization of cool-ing and heating plants operation; building management; lighting con-trol; preventative maintenance; energy auditing; and efficient bookkeep-ing.

When did centralized management of energy begin? (See Table2-1.)

As can be seen in the table, centralized monitoring and control ofequipment and conditions for HVAC systems has been around since the1950s in various forms.

In addition to reducing energy costs, the centralized monitoringand control of mechanical equipment gives an organization additionalbenefits of improved labor efficiency, reduced maintenance costs, andextended equipment life. With alarm reporting capability, mechanicalequipment problems can be noted and corrected more expeditiously.


15/244

EMS OVERVIEW 11

Table 2-1. EMS HistoryFive Generations

1st Generation (1950s)

Remote monitoring panel using temperature sensors and switches

to manually read conditions and start or stop motors.

2nd Generation (1960s)

Use of electronics, introduced low voltage circuits to automate orspeed up monitoring of panel functions.

3rd Generation (1960s-1973)Multiplexed systems consisted of groups of sensing and controlpoints tied into a local system panel and a pair of wires that run back to

a central console from multiple panels. Scanning the points in a systemwas accomplished electronically (response time was slow and failure ofthe Central Processor meant total system down).

4th Generation (1983)

Individual building panels become electronically smarter with theirown stand-alone minicomputer. They can carry out most functions that

the central computer used to do, and also relay information back to acentral console. The processing of system functions is throughout thesystem.

The speed of the electronics, as well as software, and hardware re-liability soon over powered conventional pneumatic control systems

with simple proportional control and offset. EMS sensor locations wereduplicated with pneumatic and electronic sensors.

5th Generation (1987)

Direct Digital Control (DDC) uses a small microprocessor and soft-

ware for system sensing and control. DDC units can stand alone to pro-vide various digital control sequences, or several DDC units can be tied

to a central operator station. On any size system, this could be an IBM-PC or compatible. Most EMS manufacturers have their own software

packages which results in the EMS becoming proprietary, as does the

DDC system.


16/244


FUNCTIONAL CAPABILITIES

In his zeal to conserve precious fuels and keep down growing fuelbills, the engineer often specifies the latest and most advanced EMS.What he gets is likely to be a much more complex system than is reallynecessary, one capable of performing an unneeded variety of sophisti-cated operations.

How can this be avoided, and what steps must be taken to prop-erly specify a system to assure that he gets exactly what is needednoless and no more?

Following is a brief list of events that should serve as a guide to the

overall EMS project (covered in greater detail in Chapter 11):

1. Initial Concept2. Information Retrieval3. Candidate Buildings and System Selection4. Field Survey5. Design6. Contract Documents Preparation7. Contract8. Installation and Training9. Acceptance

10. Operation and Maintenance

FUNCTIONS

The specific functions implemented in any EMS design are estab-lished by a thorough study of the building(s) and system(s) to be con-trolled. The most common EMS software functions are listed below:

Programmed Start/Stop: Occupancy schedules- Fans: save HP and heating/cooling- Pumps: can be interlocked with fans- OA Dampers: less than occupancy schedules- Air Compressors: blow down moisture

Optimized Start/Stop: Based on indoor/outdoor temperatures toachieve a comfort level. Can be stopped early.


17/244

EMS OVERVIEW 13

Temperature Setback/Setup: Change temperature set points ofthermostats when building is unoccupied.

Economizer Control: Use free cooling from outdoors when tem-perature is suitable (and) place dampers at minimum positionwhen cooling.

Enthalpy Control: Sophisticated economizer control using tem-perature and humidity (indoors and outdoors).

Discharge Air Reset: Reduce excessive heating and cooling in

HVAC systems.

Hot Water Reset: Reset hot water from outdoor air temperature.

Chilled Water Reset: Reset supply from return water temperature.

Chiller Optimization: Balance chiller operation to load demand.

Boiler Optimization: Balance boiler operation to loads and controlcombustion air.

Demand Control: Reduce peak electrical loads (kW savings).

Duty Cycling: Turn off equipment a percentage of the time accord-ing to an established schedule to reduce energy use (code compli-ance?).

Monitoring/alarm: Logging conditions, on-off/high-low alarms,trend logs over time, equipment run time, energy use, etc.

Fire Notification: Parallel with building alarm system (or) firealarm must be UL approved for this application.

Security: Alarm notification/door switches/voice synthesizers/pagers.

Card Access: Card readers, exit doors, supervised door contacts,separate programming modules.


18/244



19/244

HARDWARESYSTEM COMPONENTS 15

15

Chapter 3

HardwareSystem Components

here are five basic components that are used in energy man-agement systems. Starting from the equipment being con-trolled and working back to the operators console the compo-

nents are sensors, actuators, microprocessor-based field panels (control-lers), communication links, and a central operator station.

Sensors and/or actuators are located at the equipment being con-trolled. Sensors transmit information that defines a single operatingcondition, such as temperature or pressure. This information is supplied

to the field panels (controllers) for monitoring or decision-making pur-poses. Actuators are the mechanical interfaces that implement actionsinitiated by the controllers. The actions can be self-initiated by the con-trollers. The actions can be self-initiated or initiated as a consequence ofinformation received from the sensors. Field panels centralize the inputfrom the sensors and distribute the output from the controllers to theactuators. The information is then transmitted over the communicationlinks to a central operator station. These links carry information betweenall system components.

SENSORS

Sensors are electric devices that assess changes in ambient condi-tions and react by varying electrical voltage, or current. This voltage orcurrent variation is transmitted either as a digital or analog signal to

T


20/244


field panels, for subsequent monitoring or analysis by the controller. Adigital signal may have one of two predetermined values used to moni-tor two-position conditions, such as on/off or high/low. The analogsignal has a range of values that vary proportionally to the conditionbeing measured and is used for items such as temperature, pressure,flow, and relative humidity. Examples of analog sensors are thermo-couples, resistance temperature detectors, and thermistors. Refer to Fig-ure 3-1 for sensor mountings.

Figure 3-1. Sensor Mountings


21/244


Figure 3-1. Sensor Mountings (Continued)


22/244


Resistance Temperature Detector (RTD)The electrical resistance of certain metals varies proportionally

with temperature in a precise, consistent, and repeatable manner. RTDsmade of these metals provide a measurable resistance that is propor-tional to temperature.

Thin Film Platinum is considered the optimum sensor because ofits superior characteristics such as operating temperature range, inter-changeability, linearity, stability, and reproducibility available throughautomated manufacturing conditions. These RTDs can be furnished as 2or 3 wire, 4-20 MA units with an accuracy of 0.1% of span and a 1000ohm @ 0C reference resistance. The sensing element has a temperature

coefficient of 0.00 375 ohm/ohm/C.There are other wire wound RTDs available such as Nickel (me-

dium accuracy), Balco (low accuracy), and Nickel-Iron (70%-30%/me-dium accuracy).

All mounting configurations are available for room, duct, immer-sion, strap-on, and outside air temperature sensing.

ThermistorThermistors are a semiconductor made from combinations of

nickel, manganese, copper and other metals. They offer a fast response,are good for small spans, and are a relative low cost sensor.

Disadvantages include very non-linear, poor interchangeability,

and not suitable for wide spans. Their accuracy is 0.4F of span.

ThermocoupleTwo wires of two dissimilar metals joined to form a junction are

seldom, if ever, used with modern EMSs. They can operate over a widetemperature range however their interchangeability and repeatabilityare poor. They also have a low output sensitivity.

Relative HumidityThe principle of operation of a capacitive relative humidity sensor

is a small capacitor consisting of a hygroscopic dielectric material placedbetween a pair of electrodes. Most capacitive sensors use a plastic or onemicron thin polymer as the dielectric material, with a typical dielectricconstant ranging from 2 to 15. When no moisture is present in the sen-sor, both this constant and the sensor geometry determine the value ofthe capacitance.


23/244


By definition, relative humidity is a function of both the ambienttemperature and water vapor pressure. Therefore, there is a relationshipbetween relative humidity, the amount of moisture present in the sensor,and sensor capacitance. This relationship is at the base of the operationof a capacitive humidity instrument.

Note: All humidity sensors should be factory calibrated followingprocedures described in ASTM standard E104-85, Standard Practice forMaintaining Constant Relative Humidity by Means of Aqueous Solu-tions.

Pressure

The most common use is a pressure-electric (PE) switch where afluid pressure activates electrical contacts in the device. Differential PEswitches can be used to sense the flow of a fluid in ducts and in pipes.There are also pressure to electric transducers which will produce aproportional output electrical signal change relating to a varying inputpressure.

Devices are also used to measure static pressure in systems to con-trol fluid flow. Pressure should never exceed the calibrated range of theinstrument.

FlowThese devices measure flowrate, converting kinetic energy to a

pressure differential. Measured differential pressure typically variesfrom a few inches of water to 10 or 20 psi. See Figure 3-2 for flow sen-sors.

Accuracy or uncertainty is considered to be comprised of two com-ponents, that due to the systematize error and that due to the precisionor random error. For direct calibration of the overall meter system whichincludes the differential pressure transmitter system, the upstream anddownstream piping and suitable flow straightener, best accuracy is esti-mated to fall within the approximate range 0.2 to 0.5 percent.

A vortex shedding meter, on the industrial scene since 1970, oper-ates on the principles that the frequency of vortex shedding for fluidflow around a submersed object is proportional to the fluid stream ve-locity. Flowrate is measured by detecting this frequency. A big advan-tage for a building EMS application is that accurate measurement of theprobe output is a much simpler measurement task than accurate mea-surement of a differential pressure type meter.


24/244


Figure 3-2. Flow Sensors

A turbine meter contains a bladed rotor or turbine which rotates ata velocity proportional to volume rate of flow. Most models employ

magnetic pick-offs in which the rotor blades vary the reluctance of amagnetic circuit which generates an AC voltage in the pick-off coil. Thefrequency is directly proportional to rotor speed. This frequency issensed as an indication of flow. It can be counted by an electronic


25/244


Figure 3-2. Flow Sensors (Continued)

counter, or converted to an analog signal using voltage to frequencyconverter circuits. The calibration factor or meter factor is expressed in

electrical pulses generated per unit volume of throughput, e.g. pulses/gallon. The turbine meter has advantages of small size, repeatability, anda type of output which is digital in nature. These make the meter quitesuitable for EMS applications.


26/244


kW MeterskW Meters or watt hour meters are used to measure kW and kWh.

Input voltage must match the meter rating. Meters can be solid statewith non-resettable electromechanical display for local indication ofkWh and a contact closure for remote signaling to an EMS. Watt hourtransducers are also available which measure true watt-hours and pro-vide a pulse output to drive counters or can signal an EMS.

ACTUATORS

The actuator transforms electric- or pneumatic-coded instructionsinto mechanical responses. Actuators, which may be pneumatic, electric,electronic, or solid state, position controlled devices such as dampers orheating and cooling valves in response to signals received from thedevice controlling the actuator. Actuators may be either proportioningor modulating, with two position or snap action control.

Depending on the required control sequence of operation, if eithercontrolling air pressure or electric power is lost, actuators can fail in thenormally open or normally closed position. Pneumatic actuators re-spond to controlling air pressure changes over a range of 3 to 15 psi(pounds per square inch); electric actuators respond to on/off electricsignals from the controlling device. Proportioning electric actuators re-

spond to changes in resistance from the controlling devices, generallyover a 135 ohm range, or to changes over a 3 to 15 volt dc range for solidstate actuators.

1. Electric Relays or contractors are designed for switching electricalloads such as air conditioning, compressors, and resistance heatingappliances. Coils and contacts must be rated for the load they arecontrolling.

2. Damper Operator/Actuators open and close dampers according toan electric, electronic, or pneumatic signal from a controller. Mostoperators are spring return to normal position. Operators can beinstalled externally or internally (inside a duct).

3. Valve Operator/Actuators are the part of an automatic valve thatmoves the stem up and down based on an electric, electronic, or


27/244


pneumatic signal from a controller. For butterfly or other rotaryvalves, the operator rotates the stem. The operator and valve canbe two separate devices or together they can be one device.

4. Transducers are electro-mechanical devices that can provide elec-tric or pneumatic outputs which can be changed by the applicationof a varying electrical signal to its input.

During the 1970s, such devices were sometimes referred toas an electro-pneumatic motor driven servo. 6 or 24 volt DC powerwas used as the applied power. Positive voltage applied to theintegral motor rotated it in a clockwise direction causing an in-

crease in output pressure (negative voltage = counterclockwise =decrease in output pressure). The magnitude of the pneumaticoutput change is directly proportional to the duration of the elec-tric input signal. The output is used to reset or reposition pneu-matic controlling receivers or controlled devices.

Modern day transducers are sometimes 100% solid state us-ing a piezoresistive silicon pressure sensor and anelectropneumatic converter to provide the desired pneumatic out-put pressure. These units can be mounted in any orientation anddo not require filtered air. They provide reliable, repeatable, and anaccurate means of converting any analog signal into pneumaticpressure.

Transducers are also manufactured to provide a 4 to 20 mA or0 to 20 mA output proportional to the duration of the pulse input(pulse wave modulationPWM). Outputs may also be in the formof user selectable 0 to 10 V dc or 0 to 20 V dc depending on the enddevice requirements.

See Figure 3-3 for actuator devices.

FIELD PANELS

Field panels provide an interface between remote sensors and ac-tuators. Today these are considered to be direct digital controller (DDC)panels. Previously they were simply data gathering panels which re-layed information to a central minicomputer.

DDCs serve as a point of consolidation for many sensor and con-trol points. Each sensor or actuator represents one control point.


28/244


Uncoded signals from sensors are received, coded, and sent to the DDCmicroprocessor. Conversely, coded information is received from theDDC, decoded, and sent back to the actuators. The number of controlpoints that can be accommodated by a DDC panel varies from 4 to 200.Intelligent or smart DDCs have their own microprocessor to processinformation and respond with instructions.

Many field panels are manufactured with a built-in keyboard andvisual display. This is an item that should be specified for (at least) theprimary or master panels located throughout a facility. If the panel is notequipped with this feature, maintenance workers will have to rely on aportable operator interface device which are cumbersome and easily

left behind.Refer to Chapter 5 for additional information on digital controllers.

Figure 3-3. Actuator Devices

Pressure-electric relaysPressure-electric switches


29/244


Figure 3-3. Actuator Devices (Continued)

Valve actuators

Damper actuators

Electro-pneumatic

Motor Driven Servo


30/244


Figure 3-3. Actuator Devices (Continued)

Electro-pneumatic Transducer

Solid-state Piezoresistive Silicon

Electropneumatic Converter


31/244

SYSTEM ARCHITECTURE 27

27

Chapter 4

System Architecture

FRONT END BASED

MSs are described in this section more as a comparison as howa system of the early 1980s was designed and how it functions.

In general, these systems have field panels installed at remote lo-cations that are wired to a central computer. Field panels accept inputsfrom the remote sensors and deliver output signals to devices, however,all decisions and operating parameters reside at the central computer.For the most part, the field panels are dumb connection points. If thecentral computer is off line or if the connecting transmission wire isdisconnected all control is lost and the entire system stops functioning.

DISTRIBUTED INTELLIGENCE EMS

This type of system has a central computer and can control a largenumber of input-output points ranging from 50 to more than 2000 persystem. Field interface is provided by field panels that have limitedintelligence.

These systems are capable of performing all EMS functions, al-though, not all systems use all functions. The central computer uses

varying amounts of software storage, and tape or disc storage can beadded to increase data-handling capacity. Operator access through afixed terminal keyboard is routine. An alarm/logging printer is oftenprovided.

E


32/244


Field panels are equipped with read-only memory (ROM) in theirsoftware, which allows the panel to operate in a stand-alone limitedmode in the event of a central computer failure. Battery backup is pro-vided for the random access memory (RAM) content of the panel soft-ware and the real time clock.

For single building control, dedicated twisted pairs of wires aregenerally used for data transmission media. Where groups of buildingsare controlled, dedicated telephone lines and 1200 baud MODEMs arenormally used between buildings; with twisted pairs used within build-ings.

FULLY DISTRIBUTED EMS

The major components in a fully distributed system are stand-alone, multi-function microprocessors that have ability to perform anycombination of software functions described in Chapter 7. All necessaryapplication software is located within the microprocessor close to theequipment being controlled. The microprocessors are powered with120V and hardwired to all sensors and actuators with standard 24Vwiring. The microprocessors are looped together using any of the datatransmission methods described in Chapter 6 through either RS-232C orRS-422 communication ports. In most cases, a single communication link

ties the microprocessor loop to a central operator station. This is com-plete stand-alone operation.

There are several advantages associated with distributed systems.With the stand-alone feature, the initial investment can be limited, whileexpansion is virtually unlimited. Remote microprocessors provide localdigital and analog input/output ports, allowing direct communicationbetween the microprocessors and the sensors and actuators. Informationis transmitted in digital form. Direct digital control (DDC) pulse widthmodulation lends itself to proportional-integral-derivative (PID) control.

The central operator station that is often found with these systemsis a convenient personal computer. It is used to download applicationssoftware to remote microprocessors, edit that software, troubleshoot thesystem, and monitor or report on conditions (See Figures 4-1 through 4-5).


33/244


Figure 4-1. Andover Architecture

Andover Controls Corporation

BASIC ARCHITECTURE


34/244


Figure4-2.

Tran

eArchitecture


35/244


Figure4-3.

BarberColmanArchitecture


36/244


Figure

4-4.Johnson

ControlsArchitecture


37/244


Figure

4-5.RobertshawArchitecture

OPERATOR

WORKSTATION

DMS350

CC

HIGHWAY

D

ATA

RS23

2PORTS(2)

FANS

BOILERS

FILTERSTATUS

SMOKEDETECTOR

TEMPERATURE

HUMIDITY

PRESSURE

DAMPER

PNEU.SENSOR

VALVE

MAINLIGHTING

PARKINGLIGHTS

4CO

4CI

6AI

5PO

5PI

TOOTHERSLAVES

LIGHTING

MODULE

4CO

4CI

RS

485

SLAVE

TRUNK

DCM

6AI

5PO

5PI

TOOTH

ER

DISTRIBUTED

CONTRO

L

MODULES

(32MAX

)

DCM

6AI

4CO4CI

6AI

5PO5PI

LIGH

T-

INGMODU

LE

4CO

4CI

5PO

5PI

RS

485

SLAVE

TRUNK

TOO

THERSLAVES(UPTO4000FT.MAX.)

DCM

Distributed

Control

Module

CC

Communication

Controller

DMS350SYSTEM

ARCHITECTURE

IBM

PC

TOOTHER

DMS350s(32MAX)

PRINTER

TOOTHER

DISTRIBUTED

CONTROL

MODULES

(32MAX)


38/244



39/244

DIRECT DIGITAL CONTROL 35

35

Chapter 5

Direct Digital Control

INTRODUCTION

DC, which uses a digital computer with no moving parts, re-places both the conventional pneumatic control panel and theadded energy management system. No control devices need to

be adjusted or checked, because the microprocessor panel has no calibra-tion or routine maintenance requirements. Operating instructions builtinto the software provide for simplicity and confidence of control. Cool-ing setpoints and strategies can be set in the winter and not tested, withcomplete assurance that the DDC system will perform as expected when

summer arrives.Multiple digital control microprocessors, each operating its own

piece of HVAC or other equipment, can be linked to a single desktopconsole at a central location. Through this one desktop unit, an operatorhas access to all important setpoints and operating strategies. Monitor-ing, troubleshooting, and energy management functions are all per-formed from the same central console.

Applying a direct digital control computer to HVAC equipmentrequires only two considerations. The computer must be physically con-nected to the equipment and the computer must be given instructionsvia software on how to operate the equipment.

A DDC computer must be connected to both sensors (such as tem-

perature sensors) and controlled devices (such as valve operators). Sen-sors are connected to the computer using two kinds of inputs, analogand binary. An analog input is a variable input that could be a tempera-ture, pressure, or relative humidity reading. A binary input is a two-

D


40/244


mode input that is either on or off at any given time, such as a motorstatus, filter status, or contacts with an electrical demand meter.

Controlled devices are connected to the computer using digital andanalog outputs. A binary output is a two-mode output, either on or offat a given time. The time duration of either mode can be computercontrolled to vary between a fractional part of a second to a full on orfull off. A binary output could control a fan or pump motor or a lightingcircuit. Using pulse-width modulation, it could also control a valve ordamper actuator. Pulse-width modulation used bi-directional (open/close) pulses of varying time duration to position controlled devicesexactly as required to satisfy demand. Wide pulses are used for major

corrections, such as changes in setpoint or start-up conditions. Pulsewidth becomes progressively shorter as less correction is required toobtain the desired control setpoint.

Analog output is a variable output that might range, for example,between zero and ten volts. This is not usually needed with direct digitalcontrol because pulse-width modulation, using binary outputs, is a sim-pler and more accurate technique directly compatible with the binaryform the computer uses internally to store information.

Control of valves and dampers is very accurate with DDC becauseof proportional-integral-derivative (PID) control, perfected years ago inthe process control field. PID control techniques provide fast, responsiveoperation of a heating valve, for example, by reacting to temperature

changes in three ways: the difference between setpoint and actual tem-perature (proportional), the length of time the difference has existed (in-tegral), and the rate of temperature change (derivative) (See Figure 5-1).PID saves energy and increases accuracy at the same time by eliminatinghunting and offset by decreasing overshooting of a given temperatureand minimizing the amount of time required to settle at the desired tem-perature.

Once connections to the equipment (analog and binary inputs andoutputs) have been made, the DDC microprocessor must be given in-structions to operate the controlled devices. These instructions are in theform of software programs (application packages) with various controloptions and setpoints, all of which reside in the microprocessorsmemory.

Software, though, is what primarily determines the ultimate capa-bility of a DDC system. The changeable portions of a computer smemory provide a user flexibility of control far greater than that avail-


41/244


5-1. P.I.D. Control

INTRODUCTION

TO PID

Proportional-

Integral-Derivative

Control

One of the most

common terms

heard in connection

with todays Direct

Digital Control sys-

tems is PID; an

acronym for Propor-

tional-Integral-De-

rivative control. An

intimidating sound-

ing term, PID simply

refers to the 3 types

of control action that

are used in the con-

trol of modulating

equipment such as:

valves, dampers,

variable speed devices etc. Surprisingly the concepts behind the 3 control ac-

tions are quite straight forward and easily grasped. Gaining an understanding of

PID and its usage in control systems will provide a valuable insight into the

operation of modulating control loops.

Proportional Control- the P of PIDis a technique where a control signal

is produced based on the difference between an actual and a desired condition

(i.e. a setpoint and an actual temperature). This difference is known as the

error. The control device creates an output signal that is directly related to the

magnitude of the error, hence the name Proportional control.

Basic Proportional control is typical of that found in conventional closed

loop temperature control systems. The weakness of Proportional Control is that

it requires the existence of a significant error condition to create an output signal.

Because of this, proportional-only control can never actually achieve the desired

condition. Some small amount of error will always be present. This error is re-

ferred to as the OFFSET of the system.

Integral action is directed specifically at the elimination of Offset. Because

the magnitude of an offset is relatively small, it cannot generate a significantchange in the control signal by itself. An integrating term is used to look at how

long the error condition has existed, in effect summing the error over time. The

value produced by this summation becomes the basis for an additional control

(Continued)

PROPORTIONAL CONTROL

Figure 5-1a


42/244


signal, which is added to the signal produced by the proportional term. The result

is that the control loop continues to produce a control action over time, allowing

it to eliminate Offset.

With Proportional-Integral control we have the ability to:

1. Respond to the presence of an error in the control loop.

2. Relate the magnitude of the control signal to the magnitude of the error.

3. Respond to the existence of offset over time to achieve zero error or

setpoint.

Figure 5-1b shows the control response typically produced with Propor-tional-Integral control. The significant difference is the elimination of Offset once

the system has stabilized.

At this point one other major factor often present in modulating control

loops still needs to be addressed. That factor is Overshoot.

Overshoot refers to the tendency of a control loop to over compensate for

an error condition, resulting in a new error in the opposite direction.

As an example, consider a room with a setpoint of 72 degrees and an

actual temperature of 68. A proportional controller would respond to this error by

sending a control signal of some magnitude to the damper supplying warm air

to the room. As the room heats up the magnitude of the control signal to the

damper is reduced, but not until the room reaches setpoint would the control

signal eliminate further

heat input by closing the

damper. At this point

however the thermal in-

ertia of the room causesthe temperature to con-

tinue to rise for some

period of time. The re-

sult is that the room

overshoots the set-

point becoming warmer

than desired. The room

now requires cooling in

order to return to set-

point. The Overshoot

phenomenon not only

impacts comfort but also

results in energy waste

due to overheating andovercooling. Derivative

action (the D in PID) is

(Continued)

PROPORTIONAL INTEGRAL CONTROL

Figure 5-1b

5-1. P.I.D. Control (Continued)


43/244


designed to address

Overshoot. It provides an

anticipatory function that

exerts a braking action

on the control loop.

The Derivative term

is based on the rate of

change of the error. It

looks at how fast the ac-

tual condition is ap-

proaching the desiredcondition and produces a

control action based on

this rate of change. This

additional control action

anticipates the conver-

gence of the actual and

desired conditions, in ef-

fect counteracting the

control signal produced

by the Proportional and

Integral terms. The effect

is a significant reduction

in overshoot.

Combined, Proportional, Integral and Derivative action provide quick re-sponse to error, close adherence to setpoint, and control stability, as seen in

Figure 4. Notice the reduction in Overshoot and elimination of Offset. (Propor-

tional Integral Control Signal)

Application of PID in Building Control

While the theory behind PID control is not new there has been a dramatic

increase in its use due to the relative ease with which todays building control

systems can implement it. Once available only in expensive process control

computers, the software features of todays building control systems can provide

Proportional, Proportional Integral and Proportion-Integral-Derivative action

where needed, with relatively simple programming instructions.

The increased availability of PID control is to a large extent responsible for

the dramatic improvements in control precision seen with the use of building

control systems. Control loops such as Chiller Capacity Control, Static PressureControl. Discharge Air Temperature Control, etc. can all be controlled reliably

and precisely using PM action, providing improved operating efficiency over that

available with conventional control systems.

PROPORTIONAL INTEGRAL

DERIVATIVE CONTROL

Figure 5-1c

5-1. P.I.D. Control (Continued)


44/244


able from pneumatic control devices. This flexibility allows changingany setpoint of control strategy without interrupting system operation.DDC software, for control of HVAC and other building systems, fallsinto seven basic categories.

Sensor reading programs measure temperature, relative humidity,flow, pressure, lighting level and do other things including conversion,linearization, and square roots. They also read switch inputs (two posi-tion on/off) and totalize pulsing units (such as from power meters) tomeasure energy consumption.

On/off control programs operate start-stop devices according toanalog sensor values, such as turning on at one temperature and off at

another; switch inputs, such as manual override and device status indi-cation; and time, as in occupied and unoccupied schedules for HVACand lighting.

Modulating control programs operate variable position devices,such as valves and dampers, based on a constant, fixed setpoint, and areset schedule. An example would be resetting hot water supply tem-peratures based on the outside air temperature.

First generation DDC controllers accomplished adjust commandsusing proportional-integral-derivative (PID) control in 2 differentmethods. One method requires a feedback signal from the servo devicein order to re-adjust a control command from the PID controller. Thecontrolled variable is compared to the command or setpoint. The PID

controller then calculates how far to move from the previous spot inorder to get to setpoint. It then takes the appropriate and correctiveactions to get to that spot. Seeing how re-adjusts are done, this isclosed loop control. This method is used when programming needsthe servo feedback variable value in order to accomplish hardwaresequencing, hardware minimum positioning, etc. The second method ofcontrol is simply to make an adjustment based on the difference be-tween setpoint and variable. At predetermined time intervals the vari-able is then retested to determine the amount of change which hasoccurred. This value is then used to recalculate how much further tore-adjust to obtain setpoint. This process continues until deadband isreached. Both methods use proportional, integral, and derivative gainterms in their calculations. Both methods use PWM (Pulse WidthModulation) techniques. PWM changes its positioning device or servoby sending an output of variable time length to drive a motor whichvaries a pneumatic or electric operator.


45/244


Present-day PID controllers can use the PWM method of controlbut also can use Analog control. Analog PID controllers can be tuned foreach loops individual characteristics. The servos are integral to theanalog output, therefore, the output value is presented directly to theelectric or pneumatic operator. This value is calibrated to be in directproportion to the range of the controlled variable. For example, if thecontrolled variable is 3 degrees away from setpoint, the system knowsthat an increase of x percent will open or close the actuator the correctamount to move the controlled variable directly to setpoint. The PIDcontroller knows this because the throttling range of the actuator is cali-brated for 0.0 to 100.0 percent. PID loop tracking learns the values

needed to generate the precise output value. The value of the PID com-manded output is the actual percentage of actuator open position. Forexample, if the commanded output value is 68.0 percent, the actuatorwill be 68.0 percent open, considering there are no failures in the actua-tors themselves. This value may be read directly by the DDC for moni-toring position, or can be ranged to an Analog Data point which canchange the readout to any value however no additional hardware isrequired as in first generation DDCs with the PID feature. Changing thereadout value requires additional software points and software genera-tion. Outputs may be 0-20 mA, 0-10 vdc, or 0-20 psi, with any range ofvalues between these minimums and maximums.

Current, modern day DDC with Analog PID is by far the best

method of DDC reset control. The advantages are more accurate controlwith little or no hunting, and the elimination of the old reset servodevices which increases the chances of mechanical failures over time.

High level optimizing programs are used for pieces of equipmentwith multiple control loops and considerations, especially air handlingunit optimization (including VAV systems with or without return fantracking and guaranteeing minimum outside air ventilation) and chillerand boiler plant optimization.

Another category of programming is for energy management op-timizing routines, such as load deferral (demand limiting and duty cy-cling), optimizing start/stop; and enthalpy changeover from air han-dling units.

Alarm and reporting programs provide critical and routing alarms,data and trend logging, and energy reports.

Finally, operator interface programs can display floor plans andequipment locations, display equipment schematics and real time oper-


46/244


ating data, and provide simplified menu-driven operation.All temperature control, energy management, and automation

functions can be accomplished with these software categories.

ADJUST COMMANDS

Adjust commands on older systems were done strictly via op-erator commands from the central computer. When a command wasgiven the system would compare the difference between the com-mand and the actual position of a position of a potentiometer located

in a servo type of device. The system would then send out a voltageof the proper polarity in order to force the servo feedback to matchthe command. This was typically done on a one shot basis meaningthat if the two values did not match after one try, no other com-mands were issued automatically. The operator would be required toresend another command. The output of this servo was generallypneumatic, but in some cases was electric. The pressure output wasin no way related to the feedback readout other than by mechanicalmeans. The range of pressure output was not adjustable. If the set-point of a pneumatic controller was being reset, the only indication ofreal setpoint was the actual value of the variable being adjusted. Nocontrolled loop actions took place.

Later systems became more sophisticated in that the feedback forits adjust commands was the actual temperature itself. The operatorcommand was a temperature or humidity etc. which the system com-pared to the controlled variable for determination of how far to movethe servo device. This method was a step toward closed loop control butwas not actually because the system did no re-adjusting in order to forcethe controlled variable to the command. In this case the output pressureranges were adjustable but the output value did not reflect setpoint orposition. The controlled variable was the only indication of setpoint orposition.

ADVANTAGES

The decision to use DDC can be based on the expected value ofboth energy and labor cost savings. DDC saves significant energy dol-


47/244


lars through accurate control and by maintaining setpoint adjustmentsthat do not change with time.

Since DDC integrates temperature control and energy managementin the same system, comfort consideration can be incorporated intomore sophisticated energy management programs, such as demand lim-iting by temperature and duty cycling within deadband setpoint.

Advanced control functions are available with the microprocessor.A prime example would be calculating minimum percent outside air,using outdoor, return, and mixed air temperature sensors. Large energysavings can be realized in this way, since almost all other control sys-tems invariably use too much outside air. Once again, a small error here

produces substantial waste of heating or cooling Btus. With air volumesystems, minimum ventilation requirements can be guaranteed to pre-vent complaints resulting from stale air and improve indoor air quality.

The reliability, accuracy, and convenience of DDC reduces laborrequired for HVAC maintenance and allows for reassigning personnel toother important functions.

DDC requires both hardware and software. The hardware must bereliable, industrial grade, and engineered to interface with equipment.The software must be of a design proven to be comprehensive, flexible,and easy to use. DDC improves building operation in four ways. It re-duces energy consumption, reduces HVAC maintenance labor, improvesand assures occupant comfort, and provides greater operating conve-

nience.DDC provides enormous control flexibility and very accurate in-

formation. It allows building operators to reduce costs and providebetter services at the same time. And the life-long accuracy of DDCovercomes the inevitable decay of other controls. Computer technologyhas finally come of age in its ability to simplify and improve buildingsystems control.

The cost per point for the DDC system is usually higher than thatof the other classes, but the following additional benefits are often suf-ficient to justify the extra cost.

1. DDC systems are expandable in terms of the number of points ableto be monitored, software packages available, and operationalfunctions.

2. They are more reliable than pneumatic control systems.


48/244


3. Failure of the central operator station computer does not upset theindividual control units because satellite microprocessors are pro-grammed to stand alone in such cases.

4. Larger operator station computer memory allows building man-agement to use a preventive maintenance program and performenergy audits for the different buildings or areas of a single build-ing.

5. Electronic components are usually available from several computermanufacturers. This has the advantage that the customer is not

restricted to a particular company for equipment maintenance,and, in most cases, results in a reduction in the operation cost ofthe system.

6. Although the initial cost of DDC systems is relatively higher, thepayback period is comparable with those of smaller systems.

7. In most cases, DDCs do not reduce manpower requirements, but acentral operator-controlled system can assist in making buildingmanagement and maintenance personnel more efficient, particu-larly when implementing effective preventive maintenance pro-grams.

Microprocessors are quickly becoming a cost-effective method ofsystem control offering a superior system of distributed intelligence.They minimize host computer requirements, increase the speed andaccuracy of control, and drastically reduce system maintenance require-ments.


49/244


5-2. Pneumatic Vs DDC-D/N Stat


50/244



51/244

NETWORKING 47

47

Chapter 6

Networking

arly methods of remote monitoring or communicating be-tween two distant points relied on single pneumatic tubes orlow voltage electronic circuits to relay information. During the

1970s scanners were used with multiple systems to simultaneouslytransmit two or more messages on a single channel electronically, how-ever, response time was slow and if the central processor failed, theentire system was down.

Also during the 1970s, data communications companies, realizingthat obsolescence had overtaken conventional computer networks hav-ing one large central computer with several remote terminals connectedto it by telephone lines, determined that distributed data processing was

shown to be a more efficient way to do the job.Distributed processing, in brief, replaces the one large central com-puter with multiple smaller computers, or microprocessors, geographi-cally or functionally separated, which cooperate in the support of userrequirements. Connection between the multiple microprocessor and theother devices in the network is through various common carriers orprivate transmission methods.

Today, in the HVAC-EMS business, the transition from centralizedEMSs to distributed networks, enters the data communications world.Data Communicationsthe transmission of words or symbols from asource to a destinationis no longer exclusive to the business world. Itspervasive impact is now being felt in the HVAC-EMS profession. Under-

standing the explanations of bits, bytes, baud rate, LAN, ARCNET, to-ken ring network, IBM-PS/2, modems, RS-485, peer-to-peer, as well asa host of other communications terms will assist the EMS user to under-stand their EMS.

E


52/244


LAN (LOCAL AREA NETWORK) TOPOLOGY

During recent years, EMS manufacturers have been using explana-tions to describe their systems such as:

The System X Facility Management System utilizes a uniquetiered LAN architecture and a family of intelligent DistributedControl Units System X is based around the use of multipletoken passing LANs functioning in a tiered environment. Thisopen-ended architecture allows the system (1988-Control Sys-tems International, Carrollton, Texas).

A built-in RS-485 communication trunk is provided a secondRS-485 trunk is provided to control universal points high speed64 kilobits per second peer-to-peer LAN option enables fast sys-tem-wide response to (1989-Barber-Colman, Loves Park, Illi-nois).

The LAN selected uses industry standard ARCNET, which is atoken bus network transmitting at 2.5 Megabits to all devicescalled nodes which a second bus is based on RS-485 and Opto-22s Optomux protocol, which uses a baseband, 9600 baud, ASIIcharacter for its signaling method and (1990-Johnson Controls,Inc., Milwaukee, Wisconsin).

Network protocol is IBM SDLC operating at a speed of 1,000,000Baud communication ports consist of (2) RS232C smart control-lers with (1) optional RS232/RS485 port for (1989-Delta Con-trols, Inc., Surrey, British Columbia, Canada).

LAN, in its most basic form, is a data communication facility pro-viding high-speed switched connections between processors, peripher-als, and terminals within a single building or between buildings.

The ideal LAN would be an information distribution system thatis as easy to use as the conventional AC power distribution system in abuilding. Thus, adding a data terminal, processor, or peripheral to alocal area network should require nothing more than plugging it into aconveniently located access port. Once plugged in, it should communi-cate intelligently with any other device on the network. This ideal sys-tem is summarized by the features that make the AC power system soeasy to use:


53/244

NETWORKING 49

1. One-time installation.2. Widespread access.3. Application independence.4. Excess capacity.5. Easy maintenance and administration.

If an information distribution system were available with all thedesirable properties listed above, it would mean that telephones, dataterminals, printers, and storage devices could be moved as easily asunplugging and plugging in a lamp. Moreover, the equipment could besupplied by a variety of vendors. Although, such an ideal system does

not now exist, local area networks of several forms represent some of thefirst steps in the development of such a system.

There are four major obstacles that must be overcome in the devel-opment of the ideal LAN:

No Single StandardDue to the continually changing status of LANs and competi-

tive nature of the vendors, a variety of local area network stan-dards exist both official and de facto. The situation is improving,however, because even the dominant suppliers who have beenprotecting their proprietary interfaces are being pressured by amaturing market to release interface specifications.

Diverse RequirementsThe communications needs of a modern office building in-

clude voice, video, high-speed data, low-speed data, energy man-agement, fire alarm, security, electronic mail, etc. These systemspresent transmission requirements that vary greatly in terms ofdata rates, acceptable delivery delays, reliability requirements, anderror rate tolerance.

Costly Transmission MediaBeing able to deliver tens of megabits per second to one de-

vice and only a few bits per second to another implies that thelower rate devices are burdened with a costly transmission media.The best economic solution must involve a hierarchical networkdesign (one with stepped levels of capacity) that allows twistedpair connections for low and medium data rate devices (a low


54/244


step) feeding into a backbone high bandwidth transmission system(a higher step) such as coaxial cable or optical fibers. However, onemust be careful so that the cost of active components used forgetting on and off the network does not outweigh the lower fibercosts.

Sophisticated Functional RequirementsProviding a network with the desired data rates and dis-

tances is only one item that must be considered in the data com-munications problem. Before one data device can communicateintelligently with another, numerous higher level communications

functions must be compatible. These include codes, formats, errorcontrol, addressing, routing, flow control, access control, configu-ration management, and cost allocations.

The first, most important non-proprietary data communicationsnetwork to technically qualify as a LAN was Ethernet (a trademark ofXerox Corp.). The coaxial cable used in Ethernet can handle transmis-sions at 10 Mbs over one channel. A channel is defined as a physicallyindependent direct pathway between two devices or separate carrierfrequency on the same path.

The Ethernet architecture is based in concept on a system that al-lows multiple distributed devices to communicate with each other over

a single radio channel using a satellite as a transponder. One stationcommunicates with another by waiting until the radio channel is idle(determined by carrier sensing) and then sending a packet of data witha destination address, source address, and redundant check bits to de-tect transmission errors. All idle stations continuously monitor incomingdata and accept those packets with their address and valid check sums.Whenever a station receives a new packet, the receiving station returnsand acknowledgment to the source. If an originating station receives noacknowledgment within a specified time interval, it retransmits thepacket under the assumption that the previous packet was interferedwith by noise or by a transmission from another station at the sametime. (The latter situation is referred to as a collision, which is overcomeby networks using a baseband protocol called CSMA/CE Carrier SensedMultiple Access/Collision Detection.) The Ethernet employs the samebasic system concept using coaxial cable distribution throughout abuilding or between multiple buildings. See Figure 6-1.


55/244

NETWORKING 51

Because access to the Ethernet involves a certain amount of conten-tion (competition) between stations trying to send a message at the sametime, the behavior of the network must be analyzed and controlled in astatistical manner. Token passing networks, on the other hand, provide adifferent access procedure. Access is determined by which station hasthe token; that is, only one station at a time, the one with the token, isgiven the opportunity to seize the channel. The token is passed from oneidle station to another until a station with a pending message receivesit. After the message is sent, the token is passed to the next station. Inessence, a token passing network is a distributed polling network.

Two basic topologies (configurations or arrangements) exist for

token passing networks: Token Passing Rings and Token Passing Buses. Ina token passing ring, shown in Figure 6-2, the closed loop topology

6-1. Ethernet Configuration


56/244


defines the logical topology (that is, the order in which the token is cir-culated). A token passing bus, shown in Figure 6-3, has more operationalflexibility because the token passing order is defined by tables in eachstation. If a station (for example, a printer) never originates communica-tions, it will be a termi-nate-only station andneed not be in the poll-ing sequence. If a sta-tion needs a high prior-ity, it can appear morethan once in the polling

sequence.The forerunner of

token passing networksin the U.S. is the At-tached Resource Com-puter Network,ARCNet, developed byDatapoint Corporation.Initially, the networkand protocol were keptproprietary, but the data link protocol, interface specs, and even inte-grated circuits were made publicly available in 1982. Functionally, the

Figure 6-3. Token Passing Bus

Figure 6-2. Token Passing Ring


57/244

NETWORKING 53

Figure 6-4. ARCNET Configuration

ARCNet is a token passing bus, but the physical topology, shown in Fig-ure 6-4, is a hybrid bus/star. Rather than distribute taps along a linear busas suggested in Figure 6-3, the ARCNet uses hubs with individual portsto connect Resource Interface Modules (RIMS) to the transmission media.

The hub based architecture is an effective means of controlling thesignal quality because the hub isolates each RIM port from the maincoaxial cable. Unidirectional (one-way) amplifiers in the hubs providezero insertion loss and suppress reflections because only one direction oftransmission is enabled at a time. Amplifier switching is possible be-cause a token passing network only transmits in one direction at a time.

The ARCNet interconnects the hubs and RIMs with RG62 coaxial

cable using baseband transmission at 2.5 Mbps (baseband vs. broad-band: limits on baseband include less than 10 Mbps, low frequency,twisted pair/coax less than 2 miles; limits on broadband include greater


58/244


than 10 Mbps, digital and analog, and long distances). Although 2.5Mbps is a relatively low data rate, ARCNet uses inexpensive coax andcan be configured (laid out) with as much as four miles between sta-tions. The cable length between a hub and a RIM is limited to 2000 feet,but a four-mile span can have up to a maximum of ten hubs in a seriespath.

Local area network standards (as with other communication stan-dards) get established in two ways: by dominant manufacturers whoattract plug compatible competitors, and by official standards organiza-tions. The leading official standards organization for LANs in the U.S. isthe IEEE 802 Standards Committee. This committee has several working

groups responsible or establishing these LAN standards:

1. 802.1 Coordinating the interface between OSI Levels1 & 2 with the five higher level layers.

2. 802.2 Logical data link standard similar to HDLCand ADCCP.

3. 802.3 CSMA/CS standard similar to Ethernet.4. 802.4 Token Bus standard.5. 802.5 Token Ring standard.

Each of the LAN system architectures presented previously haveunique technical and operational advantages and disadvantages.

No presently available single LAN system architecture can eco-nomically satisfy the needs of all communications within a building orbetween multiple buildings. Nor is it likely that one system will everevolve to economically fulfill these needs. Thus, there will always be aneed for either separate systems tailored to specific applications or pos-sibly hybrid systems employing the best features of selected individualarchitectures.

EMS manufacturers that apply these hybrid-type systems will beone step ahead of the competition. (See Figure 6-5 and 6-6).

MEDIA/TRANSMISSION METHODS

Transmission MethodsGeneralA number of different transmission systems and media can be used

in an EMS for communications between the field panels and Central


59/244

NETWORKING 55

operator station. These transmission systems include twisted pairs,voice grade telephone lines, coaxial cables, electrical power lines, radiofrequency, and fiber optics. (See Figure 6-7).

Twisted pairsA twisted pair consists of two insulated conductors twisted to-

gether to minimize interference by unwanted signals.Twisted pairs can carry information over a wide range of speeds

depending on line characteristics. To maintain a particular data commu-nication rate, the line bandwidth or the signal to noise ratio may requireadjustment by conditioning the line. Twisted pairs are permanently

hardwired lines between the equipment sending and receiving data, orswitched lines routed through the telephone network. Switched lineshave signaling noise, such as ring signals within the data bandwidth,that can cause impulse noise resulting in data errors.

The nominal bandwidth of unconditioned twisted pairs is between300 and 3000 Hz. For each Hz of available bandwidth, 2 bps can betransmitted. A twisted pair with a bandwidth of 2400 Hz can support a4800 bps data rate.

Hardwired twisted pairs must be conditioned in order to obtainoperating speeds up to 9600 bps. Data transmission in twisted pairs, inmost cases, is limited to 1200 bps or less.

Voice Grade LinesVoice grade lines used for data transmission are twisted pair cir-

cuits defined as type 3002 in the Bell Telephone Company publicationstandard BSP 41004. The 3002 type line can be used for data transmis-sion up to 9600 bps with the proper line conditioning. The most com-mon voice grade line used for data communication is the unconditionedtype 3002 that allows transmission rates up to 1200 bps. Voice gradelines must be used with the same constraints and guidelines as fortwisted pairs.

Coaxial CableCoaxial cable consists of a center conductor surrounded by a

shield. The center conductor is separated from the shield by a dielectric.The shield protects against electromagnetic interference. Coaxial cablescan operate at data transmission rates in the megabits per second range.Attenuation becomes greater as the data transmission rate increases. The


60/244


Figure6-5.C.S.I.LAN


61/244

NETWORKING 57

INTRO

DUCTION

T

heSystem

7000FacilityManageme

ntSystem

utilizes

auniquetieredLocalAreaNetwork(L

AN)archi-

tecture

andafamilyofintelligentDistributedControl

Units(DCUs)toprovideacontrolnetworkofasfewas

100po

intsorasmanyas100,000points.T

hisallows

forcen

tralizedcommandandcontrolofmanywidely

distribu

tedprocesseswithunprecedentedreliabilityand

speed.

OVERVIEW

T

heSystem

7000isbasedaroundtheuseof

multipletokenpassingLANsfunctioninginatiered

environ

ment.Thisopenendedarchitecture

allowsthe

system

tosupporthundredsofterminalworkstations

andtensofthousandsofpointswithrespo

nsetimes

notfou

ndintodayssystems.

HOSTLAN

AtthetopofthistieredLANarchitectureisone

ormoreHostLANs.As

inglehostLANcansupporta

singlePC

workstation

orasmanyaseight(8)PC

workstations.Allwork

stationsarecommunicating

overasinglepairofwire

sontheHostLANinamulti-

tasking/multi-userenviro

nmentwithoneormorecon-

trollerLANs.

UtilizingtheHost

LAN

structure,oneormore

workstationsmaycommunicatewithover1,000re-

motecontrollerLANs,e

achcapableofsupporting63

DistributedControlUnits.

TheHostLAN,aswellasthecontrollerLAN

utilizesaCSIproprietar

yprotocolforLANcommuni-

cations.Equipmentnot

designedfordirectcommuni-

cationwiththeLANmu

stutilizeagatewayorTAP

intotheLAN.ThesefirmwarespecificTAPsprovide

formessaging,protocol

conversion,AA/AD(AutoAn-

swer/AutoDial)andmu

chmore.


62/244


INDUSTRY STANDARD ARCNET

ARCNET is a token bus network Initially developed by Datapoint Corporation as a

very fast, 2.5 Megabit communications link for computer-to-computer connections. The

technology was licensed to third party electronics firms to manufacture ARCNET control

chips, making low cost token passing networks readily available to many industries. Itwasnt long before ARCNET moved from the office to the shop, as system integrators

realized that the networks reliability, noise immunity, predictability, and low cost were

perfect for allowing automated machines to communicate with each other on the factory

floor. Growth in demand led to second sourcing of the control chips, and the development

of new configuration and cabling options. As many more manufacturers adopted the use

Figure 6-6. J.C.I.LAN


63/244

NETWORKING 59

of ARCNET in their systems, the ARCNET Trade Association was formed to provide a user

forum for maintaining standards, coordinating connectivity issues, and charting future

growth. ARCNET has now become a de facto industry standard, with over 1,000,000

connected devices worldwide.

TOKEN BUS

In ARCNET, devices connected to the network are called nodes, which are ad-

dressed from 1 to 255. Access to the network is controlled by a token which is passed

around the network, going from each node to the node with the next address. When anode has the token, it may broadcast a message to any other node before passing the

token on. This message is received simultaneously by all other nodes, but only responded

to by the node to which the message was addressed. This scheme allows all devices on

the network to operate on a peer-to-peer basis, which means the network is not dependent

on any single device for nodes to share information. Token passing is also deterministic,

which means that the maximum amount of time it takes for a message to be sent from one

node to another is predictable, even under heavy communication traffic conditions. It also

guarantees that every node has access to the network on an equal basis.

ARCNET is self-configuring. If a node should fail, it is automatically removed from

the token passing sequence so that communication is maintained uninterrupted among the

remaining nodes. When a new node is added, or a failed node recovers, ARCNET imme-

diately recognizes the node and adds it into the token passing sequence. Should the

communication trunk be severed, both halves of the network are automatically

reconfigured Into two separate networks, each with the ability to maintain peer-to-peer

communications among the connected nodes.Network Control Units and Operator Workstations are ARCNET nodes in Metasys.

NCUs optionally have an ARCNET communication circuit integrated within the Network

Control Module. For an Operator Workstation, an ARCNET communication card is installed

in a PC to allow the computer access to the network. ARCNET cards are manufactured

by both Johnson Controls and other vendors.

Figure 6-6. J.C.I.LAN (Continued)


64/244


Figure 6-7. Transmission Media

HIGH DATA RATE ISOLATION

IMMUNITY TO NOISE SMALL SIZE, WEIGHT

OPTICAL FIBER

ELECTRICAL

SIGNAL

CLADDING

CORE

ELECTRICAL

SIGNAL

LIGHT

SOURCE

LIGHT

DETECTOR

Outter conducter braided shield

Inner conductor solid metal

Separated by insulating material

Covered by padding (some does not meet fire code)

INSULATION

INNER CONDUCTOR

OUTER COATING

COAXIAL CABLE

COPPER WIRE PAIRS

Separately insulated

Twisted together Usually installed in buildings when built

Often bundled into cables Bandwidth limited to approximately 50 kHz

1 twisted pair = 2-wire circuit = 1 local loop

2 twisted pairs

= 4-wire ckt.


65/244

NETWORKING 61

transmission rates are limited by the data transmission equipment andnot by the cable. Regenerative repeaters are required at specific intervalsdepending on the data rate, nominally every 2000 feet to maintain thesignal at usable levels.

Power LinesData can be transmitted to remote locations over electric power

system lines using carrier current transmission that superimposes a lowpower RF signal, typically 100 kHz, onto the 60 Hz power distributionsystem. Since the RF carrier signal cannot operate across transformers,all communicating devices must be connected to the same power circuit

(same transformer secondary and phase) unless RF couplers are in-stalled across transformers and phases permitting the transmitters andreceivers to be connected over a wider area of the power system.

RFModulated RF can be used as a data transmission method with the

installation of radio receivers and transmitters. The use of RF must becoordinated with the communications department to avoid interferencewith other facility RF systems. MODEMS must be provided at each re-ceiver-transmitter location. FM is used in most cases because it is notsusceptible to amplitude related interference.

Fiber OpticsFiber optics uses the wideband properties of infrared light travel-

ing through transparent fibers. Fiber optics is a reliable communicationsmedia which is rapidly becoming cost competitive when compared toother high speed transmission methods. It is best suited for point-to-point high speed data transmission.

The bandwidth of this media is virtually unlimited, and extremelyhigh data transmission rates can be obtained. The signal attenuation ofhigh quality fiber optic cable is far lower than the best coaxial cables.Repeaters required nominally every 2000 feet for coaxial cable, are 3 to6 miles apart in fiber optic systems. Fiber optics must be carefully in-stalled and cannot be bent at 90 right angles. Additional benefits in-clude features such as space savings in conduits and freedom from EMIinterference. However, on the other hand, splicing is difficult and thereis the requirement of convertors to get off the fiber optic network.

See Table 6-1 for a comparison of transmission methods.


66/244


Table6-1.Tran

smissionsMethodsComparison

Compatibility

First

Scan

Maint.

WithFuture

Method

Cost

Rates

Reliability

Effort

Expandability

Requirements

Coaxial

high

fast

excellent

min.

unlimited

unlimited

Twisted

pair

high

med.

verygood

min.

unlimited

limited

RF

med.

fastbut

low

high

verylimited

verylimited

limited

Microwa

ve

veryhigh

veryfast

excellent

high

unlimited

unlimited

Telephon

e

verylow

slow

lowtohigh

min.

limited

limited

Fiberop

tics

high

veryfast

excellent

min.

unlimited

unlimited


67/244

NETWORKING 63

MODEMS

MODEM is an acronym for modulator/demodulator. The MO-DEM is analogous to a telephone set, which converts the voice to anelectric signal at one end of a wire and converts the signal back to soundat the other end of the wire. MODEMs can communicate between fieldpanels and controllers when the controller is separate from the fieldpanel. MODEMS are also used to communicate with an EMS from aremote location. Upon receiving a signal from a sensor or controller, theMODEM imposes the information in binary form onto carrier waves.These waves convey information over communication links known as

data transmission media. The information is imposed on the wave byaltering, or modulating, the wave form; it is then extracted from thewave by demodulating. In the case of a digital signal from a sensor thisprocess is straightforward. Analog signals from sensors require analog/digital converters to condition the signal prior to modulation, MODEMsare characterized by transmission speed and the way in which modula-tion is accomplished.

There are two basic modulation classificationsbaseband andbroadband. Baseband MODEMs convert data into binary form usingdifferential current impulses for transmission. However, baseband is nottrue modulation because a carrier wave is no