/5;2 :2
`
`{R i c Eqngsmifis
`
`AND ZONED HEATING
`
`MODULATING FURNACE a, I
`SYSTEM DEVELOPMENT aVI
`
`FINAL REPORT
`July 1987 — December 1989
`
`
`
`fil I
`firl
`
`
`
`GasResearchinstitute
`arl
`
`8600 West Bryn Mawr Avenue
`Chicago, Illinois 60631
`
`
`
`HONEYWELL — EXHIBIT 1002
`
`HONEYWELL - EXHIBIT 1002
`
`

`

`Report No. Til-150901691
`
`GRI-91/0075
`
`R E C E I V i"; :3
`
`AUG : ,9 1391
`DEVIYICE OF
`339""“"" A mnAQwYRATW‘":
`
`MODULATING FURNACE AND ZONED HEATING SYSTEM DEVELOPMENT
`
`FINAL REPORT
`July 1987 - December 1989
`
`Prepued by:
`8. J. Feldman. at 1:1.
`
`Tecogen Inc.
`45 First Avenue
`P.O. Box 9046
`Waltham. Massachusetts 02254—9046
`
`Prepared for:
`
`GAS RESEARCH INSTITUTE
`Contract No. Gum-50874454522
`
`Building System- Research
`
`January 1991
`
`

`

`Gm DISCLAIMER
`
`LEGAL NOTICE —~ This report was prepared by Tecogen Inc.. a subsidiary of
`Thermo Electron Corporation. as an account ofwork sponsored by the
`Gas Research Institute [GRIL Neither GR]. members of GRI. nor any person acting
`on behalf of either:
`
`:1. Makes any warranty or representation, express or implied. with respect
`to the accuracy. completeness. or usefulness ofthe information
`contained in this report. or that the use of any information. apparatus.
`method. or process disclosed in this report may not infringe privately
`owned rights: or
`
`b. Assumes any liability with respect to the use of. or for damages resulting
`from the use of. any information. apparatus. method, or process
`disclosed in this report.
`
`

`

`lillllllllllllllllllIiIiiiiiiiiiliiiiiiiliilllllllllllll
`
`I. Dowel D!!!
`January 1991
`
`C.
`
`502724]
`L heron! no.
`REPORT DOCUMENTAHON
`
`
`GRl-91/0075
`PAGE
`
`
`4. 7’“. Ind Whittle
`
`
`
`Modulating Furnace and Zoned Heating System Development
`1. MW“)
`
`l. Foaming Organisation loot. Mo.
`
`S.J. Feldman et al.
`TRQSOB’OlG-S‘l
`9. Performing Organisation Home Ind Address
`
`IO. WIT-“Moth UM "C.
`
`
`
`Tecogen Inc.
`u. tantrum?) or ammo) Ne.
`45 First Avenue
`
`P.0. Box 9046
`(Cl5087’2hs-1522
`
`02254‘9046
`
`
`Waltham, Massachusetts
`
`“D
`
`
`11. Sponsoring Org-Mullen Name Ind Adams
`It. Type 00 Report I. foiled Covered
`
`
`Gas Research Institute
`Final Report
`
`
`
`' 8600 West Bryn Mawr Avenue
`Jul
`1987 - December 1989
`
`60631
`Chicago, lllinois
`
`
`ll. Iupplememuy "do:
`
`
`
`
`«1.. Matteo! (Unit: 200 W6!)
`
`
`
` This report describes an experimental modulating furnace and a zoned warm air heating system for use
`
`in residences.
`The system was installed and tested at the National Association of Home Builders'
`(NAHB) SMART HOUSE in Bowie, Maryland.
`The key features of this system include:
`(1) continuous
`
`
`
`
`modulation of firing rate and supply air over a wide range,
`(2) closed-loop control
`to maintain a
`
`
`desired supply air flow under varying system resistances,
`(3) continuous modulation of combustion
`
`
`air to maintain efficiency,
`(4) a proportional-integral control algorithm operating on measured
`
`
`
`temperatures and set points in each zone to set the furnace firing rate,
`(5)
`low-cost on/off dampers
`to direct air flow to those zones calling for heat, and (6) a single microprocessor—based controller
`
`
`that integrates all aspects of the system.
`
`17. Document Mow.
`
`a. Outflow“
`
`Residential warm-air heating systems
`Residential zone heating
`Residential modulating furnace
`
`I. Milieu/MW“ You!“
`
`:. COM“ Field/EMF
`
`m. nun-mm 3:an
`
`UM WWW
`(5.. mat-nus)
`
`
`no. ”out; cu... mm upon)
`2:. No. 0' n...
`Unclassified
`“6
`a. Security cl.” rm. no) P
`8“ Instructions on nmm
`omen». mm 212 u—m
`(formerly HHS-)5)
`Department of Commute.
`
`

`

`RESEARCH SUM'MARY
`
`TITLE
`
`Modulating Furnace and Zoned Heating System
`Development
`
`CONTRACTOR
`
`Tecogen Inc.
`GRI Contract Number: 5087445-1522
`
`PRINCIPAL
`INVESTIGATORS
`
`S.J. Feldman. D. Bartz. E. Doyle. J. Tandler.
`andA Lowenstein
`
`REPORT PERIOD
`
`July 1987 — December 1989
`Final Report
`
`OBJECTIVES
`
`TECHNICAL
`PERSPECTIVE
`
`RESULTS
`
`Development of a continuously modulating residential gas
`furnace and a zoned heating system.
`
`GRI's support for the development of the modulating furnace
`will accelerate the introduction of this needed product into
`the marketplace. This product will bring to gas consumers
`unprecedented levels of comfort and convenience.
`It‘s the
`next logical step in the evolution of space conditioning.
`When coupled with zoning. the modulating furnace can offer
`significant energy savings.
`
`Some of the benefits provided by a modulating furnace
`include: (I) reductions in electrical consumption,
`(2) reductions in cycling frequency. (3) improvements in
`comfort due to greater temperature uniformity.
`(4) reductions in noise. (5) an improvement in seasonal
`efficiency. and (6) efficient operation of the furnace in an
`arrangement where many small zones need to be serviced.
`
`Fumace manufacturers need a new top-of-the-line product.
`The modulating furnace is the ideal candidate. but before
`they can achieve this goal. outstanding development issues
`must be solved.
`
`Tecogen developed for GRI an experimental modulating
`furnace and a zoned warm air heating system for use in
`residences. The key features of this system include:
`(1) continuous modulation of firing rate and supply air over
`a wide range. (2) closed-loop control to maintain a desired
`supply air flow under varying system resistances,
`(3) continuous modulation of combustion air to maintain
`efficiency. (4) a proportional-integral control algorithm
`operating on measured temperatures and set points in each
`zone to set the furnace firing rate. (5) low-cost on/off
`dampers to direct air flow to those zones calling for heat. and
`(6) a single microprocessor—based controller that integrates
`all aspects of the system.
`
`

`

`
`
`modulation and zoning and the benefits they oifer. Testing
`provided valuable data on different control algorithms and
`brought to the fore the outstanding technical issues
`
`TECHNICAL
`APPROACH
`
`PROJECT
`IMPLICATIONS
`
`The approach taken in developing the modulating furnace
`recognized that commercialization would be difficult if
`significant changes were made to the furnace's basic design.
`So the emphasis was on developing a retrofittable "controls
`package" for converting fixed-firing rate furnaces to
`modulation. The goal was to develop an experimental
`version of this controls package and to show how it works in
`the field.
`
`This project was supported by GRI to develop a new
`residential heating system that has the comfort benefits
`associated with modulation and the energy savings potential
`associated with zoning. The zoned wann-air heating system
`
`This project produced results on the performance of the
`furnace under a variety of operating conditions in the
`laboratory and in a research house with a zoned warm-air
`distribution system. Future research is planned to refine
`the control system and evaluate commercial opportunities.
`
`Larry Brand
`GRI Project Manager
`
`

`

`TABLE OF CONTENTS
`
`RESEARCH SUMMARY .......................................
`
`1. EXECUTIVE SUMMARY ...................................
`
`l. 1 OVERVIEW .........................................
`
`1.2 RA'IIONAIE .........................................
`
`1.3 ACCOMPLISHMENTS ..................................
`
`1.4 RECOMMENDATIONS .................................
`
`2. WORK PERFORMED ......................................
`
`2.1
`
`INTRODUCTION .......................................
`2.1.1 Program Scope .................................
`2. 1.2 Technical Approach .............................
`2.1.3 Elements of the SMART HOUSE System ..............
`
`2.2 SMAR’I‘ HOUSE SYSTEM DESCRIPTION ...................
`2.2.1 Furnace ......................................
`2.2.2 Gas Control ...................................
`2.2.3 Circulating Air Control ...........................
`2.2.4 Combustion Air Control ..........................
`2.2.5 Hard-Wired Control Logic .........................
`2.2.6 Single-Board Controller ..........................
`2.2.7 Data Acquisition and Monitoring ...................
`2.2.8 Zone Control ..................................
`2.2.9
`SMARI‘ HOUSE ................................
`
`2.3 FINDINGS FROM SMART HOUSE TESTS ...................
`2.3.1
`Single-Zone Operation ...........................
`2.3.2 Multi-Zone Operation ............................
`
`2.4 SEASONAL PERFORMANCE ANALYSIS ....................
`2.4.1 Results .......................................
`2.4.2 Definition of the Load ...........................
`2.4.3 Definition of the Conventional Furnace ..............
`2.4.4 Definition of the Modulating Furnace ................
`
`v
`
`1
`
`1
`
`2
`
`3
`
`3
`
`7
`
`7
`7
`7
`10
`
`10
`10
`13
`13
`15
`15
`19
`2 1
`2 1
`23
`
`23
`23
`33
`
`38
`38
`41
`41
`43
`
`

`

`LIST OF ILLUSTRATIONS
`
`2.1
`
`2.2
`
`2.3
`
`2.4
`
`2.5
`
`2.6
`
`2.7
`
`2.8
`
`2.9
`
`2.10
`
`2.11
`
`2.12
`
`2.13
`
`2.14
`
`2.15
`
`2.16
`
`2.17
`
`2.18
`
`2.19
`
`2.20
`
`2.21
`
`Test Facility .........
`
`................................
`
`Typical Operating Envelope for a Furnace Operating
`................................
`at a Fixed Air Flow Rate
`
`Elements of the SMAKI‘ HOUSE System ....................
`
`Furnace Used at the SMART HOUSE ......................
`
`Desired and Actual Gas Control During a Typical Period
`................................
`at the SMART HOUSE .
`.
`
`Desired and Actual Circulating Air Flow Rate
`During a Typical Period at the SMART HOUSE ...............
`
`Circulating Air Fiow Rate Control Schedule ..................
`
`Factory—Wired Control Logic .............................
`
`Control Logic Used with Microprocessor ....................
`
`Single-Zone Control: Typical Day ........ . ................
`
`Single-Zone Control: Low Load ................... . .......
`
`Single-Zone Control: High Load ..........................
`
`Single-Zone Control: Qmt Based on Historical Load ...........
`
`Single-Zone Control: High Gains .........................
`
`Mum-Zone Control: Zone Units Proportioncd to Supply Registers
`
`Multi—Zone Control: Zone Temperatures ....................
`
`Mum—Zone Control: ZU2 = 4-0 Percent .....................
`
`Mum-Zone Control: Basement Unconditioned ...............
`
`Seasonal Heating Load Profile ............................
`
`Standard Thermostat Cycling Rate ........................
`................................
`
`Strategy for Modulation
`
`ll
`
`12
`
`14
`
`16
`
`l7
`
`18
`
`20
`
`27
`
`28
`
`29
`
`30
`
`32
`
`34
`
`35
`
`36
`
`37
`
`42
`
`45
`
`

`

`LIST OF TABLES
`
`2.1
`
`2.2
`
`2.3
`
`2.4
`
`Values of Zone Control Parameters ........................
`
`SMART HOUSE Zone Description .........................
`
`Seasonal Energy Consumption ...........................
`
`Seasonal Operating Costs ...............................
`
`24
`
`25
`
`39
`
`

`

`1. EXECUTIVE SUMIMARY
`
`1.1
`
`OVERVIEW
`
`Tecogen developed for GRI an experimental modulating furnace and a zoned
`warm air heating system for use in residences. The key features of this system
`include:
`i 1) continuous modulation of firing rate and supply air over a wide range.
`
`(2) closed-loop control to maintain a desired supply air flow under varying system
`resistances (3) continuous modulation of combustion air to maintain efficiency.
`(4) a proportional-integral control algorithm operating on measured temperatures
`and set points in each zone to set the furnace firing rate. (5) low-cost on/ off
`dampers to direct air flow to those zones calling for heat. and (6) a single
`microprocessor-based controller that integrates all aspects of the system.
`
`Rather than designing from scratch a furnace or a burner optimized for
`modulation. the approach taken was to develop a factory installable "controls
`package" for converting existing fixed-firing rate furnaces to modulation. Starting
`from scratch would make commercialization difficult because it would force
`
`manufacturers to invest heavily to retool for the new design. The goal was to
`attract a furnace manufacturer by developing an experimental version of the
`controls package and demonstrating it in the field. The field test site was the
`National Association of Home Builders' (NAHB) SMART HOUSE in Bowie. Maryland.
`
`Development of the modulating furnace was limited to its application to
`high-efficiency condensing furnaces.
`It was felt that the modulating furnace would
`initially be marketed as a top«of—the—line product and as such would have to have
`as a base system the most efficient furnace available.
`
`Pulse combustion systems are an important element of the condensing
`furnace market but were outside the scope of this prog‘am.
`it may be possible to
`modulate pulse combustion systems. but this program focused exclusively on
`induced draft systems since these Were the prevalent designs at the time of the
`program.
`
`Zone temperature control at the SMART HOUSE was accomplished using a
`combination of feedback and feedforward control. The feedforward algorithm
`estimates the house load based on either historical energy use (a time average of
`the firing rate over the previous two hours): or outside air temperature, zone
`setpoints. and an assumed UA Both methods were tested. The feedback algorithm
`corrects the estimate by comparing current temperatures with setpoints. The result
`of this computation is a "desired" furnace firing rate that should very closely match
`the load.
`
`

`

`
`
`if it's underflmig or cycles more often if it's overm-ing. but since combustion air is
`also open loop. proper stoichiometry is not guaranteed. fins is an area needing
`further attention.
`
`Combined gas and electric operating cost savings of about 8 percent are
`projected for the modulating furnace operating in a single-zone system. (Based on
`Boston weather data.) With zoning. even larger savIngs are possible.
`
`

`

`1.3
`
`ACCOWLISHMENTS
`
`'l‘ecogen built two modulating furnaces with zone control under 6121
`sponsorship. One was installed at the Natioi'ial Association of Home Builders'
`(MAI-IE) SMART HOUSE in Bowie. Maryland. and the other in Teoogen's laboratory.
`The SMART HOUSE unit operated reliably for 3 months during the 1989 winter.
`These two units were experimental and were designed to evaluate the modulation
`and zoning and the benefits they offer. While they are not commercially viable
`systems in their present form. testing provided valuable data on diiferent control
`algorithms and brought to the fore the many outstanding technical issues.
`
`1 .4
`
`RECOHIMENDATIONS
`
`The experimental system built by Tecogen for GRI demonstrated many of the
`features of a modulating furnace. However. several areas of development are still
`outstanding and must be addressed before a modulating furnace can be
`commercialized.
`
`Without doubt the most important outstanding technical issue is the need
`for a method to control stoichiometry or to keep stoichiornetry within a safe range
`to assure that the CO limit of 400 ppm is not exceeded while modulating. Also. the
`limit must not be exceeded if the flue should be partially or fully blocked. as
`prescribed by ANSI 221.64. The assumption is that one wants to operate as close
`to the CO limit as possible to maintain high thermal efl'iciency.
`
`The present method of using open loop control for both gas and combustion
`air is unacceptable unless appropriate reliable safety switches are incorporated into
`the system. Possible failure paths with the present system include:
`
`(2) Gas valve modulates.
`(1) Gas valve fails to modulate. i.e.. sticks open.
`but to an unpredictable pressure. This may happen because of manufacturing
`tolerances. changes in valve performance over time. or changes in valve
`performance resulting from changes in ambient conditions.
`(3) Gas valve
`modulates to the right pressure. but changes in gas composition or density alters
`gas mass flow. air-fuel ratio and combustion process.
`
`If a reliable low-cost CO "switch" were available. then the present system
`would be acceptable. except a new modulating gas valve needs to be developed.
`The valve should combine the features of the standard combination gas control
`valve and the modulating/regulating valve used at the SMART HOUSE. The key
`technical specification is a 4:1 turndown or greater. High turndown allows
`servicing of small zones. greater numbers of zones. and operation with fewer
`furnace cycles. Existing modulating/regulating valves have only a 2.0 to 2.6:]
`turndown.
`
`

`

`
`
`reliability for this application.
`
`Developing an appropriate method for stoichiometry control must also be
`coordinated with those responsible for developing safety standards. Standard-
`
`The present system uses a proportional-integral (PI) based algorithm with
`a feed-forward estimate of load to set the firing rate. This approach worked well at
`the SMART HOUSE, but further testing is needed. The proposed algorithm was
`tested for one season in one house. A much broader sample tested under a variety
`of situations — different loads. housing types. climates. usage patterns. etc. - is
`needed. Also. several changes need to be incorporated into the algorithm based on
`lessons learned from the SMART HOUSE tests.
`
`

`

`thermostat? Should the thermostats simply be l/O-sensor terminals networked to
`a central processor in the furnace? This would lower the cost per thermostat.
`which is critical for stimulating a market for zoning. Or. should each thermostat
`have processor capabilities on its own? This allows modulating thermostats to, be
`sold separately from modulating furnaces. in the same way that thermostats and
`furnaces are now sold.
`
`is it worth the cost of making the thermostats a two-wire system? Doing so
`would lower installation cost in retrofit situations because all present heating
`systems have at least two wires. Also. should the furnace and thermostat use a
`communication protocol compatible with CEBus or SMART HOUSE? Such systems
`may be the "wave-of-the future."
`
`The thermostat must be capable of operating the air conditioner. How will
`this affect the design of the thermostat? Must one thermostat be capable of
`operating both variable-capacity and smgle-capacity air conditioners? This project
`did not address air conditioning needs.
`
`

`

`

`

`2. WORK PERFORMED
`
`2. 1
`
`INTRODUCTION
`
`2-1-1W
`
`The goal of this project was to develop a continuously modulating residential
`gas furnace and zoning system.
`
`Development of the modulating technolog was limited to its application on
`high-efficiency condensing furnaces. This type of furnace was chosen because it
`is the type most amenable to conversion for modulation.
`
`2. 1.2
`
`e hnical
`
`roach
`
`Furnace design is governed by three independent parameters: firing rate.
`circulating air flow rate. and combustion air flow rate. The first step in the program‘
`was to build a test rig (Figure 2.1) where the efl‘ects of independently varying these
`parameters could be measured.
`in so doing it was possible to define an operating
`envelope for the furnace. The boundaries of this envelope are governed by design
`limits such as the point where condensate forms in the primary heat acchanger. the
`point where excessive heat exchanger temperatures are reached. the point where
`too hot or too cold supply air temperatures are reached, and the point where
`excessive carbon monoxide is generated.
`
`Figure 2.2 shows one such operating envelope. The data were taken at a
`fixed circulating air flow rate. This particular furnace's design point is shown by
`the square in the figure. The effects of moving ofi' the design point are clearly seen.
`For example. the condensation limit is reached at or about 68 MBtu/hr if excess
`air and circulating air flow rates remain constant while firing rate is decreased.
`Avoiding this limit while reducing firing rate even further requires increasing access
`air.
`if this is done efilciency decreases somewhat. Maps like this were used to aid
`in developing control strategies for the modulating furnace.
`
`Zone temperature control at the SMART HOUSE was accomplished using a
`combination of feedback and feedforward control loops. The feedforward algorithm
`calculates the nominal house load based on either historical energy use or outside
`air temperature and zone setpoints. Both methods were tried. The feedback
`algorithm calculates a correction to this nominal load based on a comparison of
`current temperatures and setpolnts.
`
`

`

`
`
`DISCHARGE
`
`COOLING
`
`167-188r
`
`CEILING
`
`BLOWER
`
`TEST FURNACE
`
`BOOSTER
`
`FLOOR
`
`Figure 2.1 Test Facility
`
`

`

`meumhu...
`
`Egg/mum,,r,
`
`onecomcomon“
`
`ANVmH<mmmuxw
`
`>m<szm2Honp<mztzou
`
`./
`
`memmEéEzEmém/%\\\K1/;1/I;a;/\f\\\
`
`III(”4‘O‘V\’D
`
`
`
`Szszw/mow,4%,
`
`III
`
`am
`
`0v
`
`om
`
`om
`
`(spuosnoql)
`
`(H/nuaw> 31V8 SNIHIJ
`
`
`
`3mm30EL_<68:".manmczmhmao085:“.ahemao_m>cwmczeao.333.~.~83mm“.
`
`
`
`
`
`
`
`
`

`

`
`
`analysis. The microprocessor controlled the inducer motor. circulating air motor
`
`modulating furnace would couple to a zoned system. and (3) sewe as platform for
`testing hardware and various furnace and zone control algorithms.
`
`2.2
`
`SMART HOUSE SYSTEM DESCRIPTION
`
`2.2. 1 Furnace
`
`The furnace tested at the SMART HOUSE was a modified version ofa
`commercial condensing furnace with a rated input of 100 MBtu/hr and an AFUE
`of 90.5 percent.
`it was designed to operate with a temperature rise between 50°
`and 80°F. This implies an air flow rate between 1
`
`anger cells the test furnace actually had
`foul-J The furnace uses aluminized steel multi-port burners which fire up into
`glass~1med primary heat exchanger cells. Combustion gm pass through the cells
`
`secondary heat exchanger assembly. This assembly consists of three staggered
`rows of half-inch diameter stainless-steel tu
`enhance heat transfer. Aluminum fins are attached to these tubes to enhance heat
`transfer to the circulating air.
`
`10
`
`

`

`5N3“:
`
`mmEmeD
`
`hzmimm<m
`
`l'lll'l'lll'l'lII-IIII
`D
`
`
`
`
`
`239$mmno:5.5%.2:ho3:95522:9”.
`
`‘quo 30'"!
`
`na-
`
`$40
`
`ll
`
`wZON
`
`mmmDh<mwn3mh
`
`
`

`

`
`
`PRIMARY
`HEAT
`EXCHANGER
`
`71-168
`
`"is INTERCONNECTING
`PIPE
`
`SECONDARY
`HEAT
`EXCHANGER
`
`" CIRCULATING
`AIR BLOWER
`
`
`
`
`
`INDUCER'
`BLOWER
`
`=
`2':
`
`"
`
`BURNERS
`
`GAS
`VALVE .
`
`s
`CONTROL "
`MODULE ;
`
`Figure 2.4 Furnace Used at
`
`the SMART HOUSE
`
`12
`
`

`

`The multi-port burners were easily modulated. By adjusting excess air
`levels. a stable blue flame was maintained at inputs less than 20 MBtuh/hr. over
`a five-to-one turndown.
`
`This furnace is not a direct vent type; combustion air is drawn from the
`space around the furnace. not ducted from outside. So the unmodified unit was
`certified per ANSI 22 1.47. American National S tandarclfor Gas-Fired Central
`Fumaoes {Except Direct Vent and Separated Combustion System Central Muses).
`Most furnaces in the future will be the direct-vent type and will be certified under
`ANSI standard 221 .64.
`
`2.2.2 Gee Control
`
`An experimental modulating]regulating valve made by Maidtrol Corporation
`(Model E5281) controlled firing rate.
`it was placed in series with and downstream
`from the Honeywell CGC (combination gas control) valve that comes standard with
`the furnace. This configuration made it necessary to slightly increase the 060's
`normal setpoint of 3.5 MG [inches water column. gauge) to 3.8 M0 to overcome
`the added pressure drop of the Maxitrol valve. The Maxitrol valve regulated
`manifold pressure between 3.5 and 0.32 IWG. Pressure regulation was controlled
`by voltage to the valve. Higher voltages resulted in higher pressures. Power to the
`valve was applied in a pulse-width-modulated fashion because this reduced
`hysteresis.
`
`Data in Figure 2.5. which covers a ten-hour period on a typical day at the
`SMART HOUSE. shows valve performance to be excellent. The dotted line shows
`
`the desired firing rate. The actual firing rate. as measured by a mass flow meter.
`is shown by the solid line. From this we see the valve is fast-responding and
`accurate over the full modulating range of 100 to 30 MBtu/ hr.
`
`2.2.3 Circulating Air Control
`
`'l‘ecogen replaced the standard circulating air blower motor in the furnace
`with a General Electric brushless DC motor. An electronics module, also provided
`
`by GE. controlled the motor's speed using electrical commutation. The
`motor/ controller package. called an ECM (electrically commutated motor). has a
`better cost/benefit ratio when compared with alternative methods of speed control
`for this application such as inverter-drives or SCR‘s.
`
`Tecogen developed a method to achieve wide-ranging closed-le control of
`air flow under varying system resistances using the controller’s tachometer output.
`This approach made it unnecessary to use costly temperature or pressure
`transducers to achieve the same end.
`
`13
`
`

`

`EN-3E.
`
`5:
`
`«a2.ooe
`
`@565us:
`
`
`
`Gum—mun........
`
`
`
`-_<=._.0<1'
`
`(an/mew) suvu smau
`
`14
`
`
`
`
`
`
`
`
`
`wmno:FMS;9:«mDatum:83»...m9.1.5.9550m5_m30<ucmUmtmoam.~9.sz
`
`
`
`

`

`Closed-loop control of air flow at the SMARI‘ HOUSE is shown in Figure 2.6.
`The graph shows a typical 19-hour period during which the desired flow rate.
`shown as the dotted line. varied from 1300 cfm to 845 cfm. The actual flow rate.
`as measured by a velocity probe. is shown as a solid line and follows closely the
`desired value.
`
`For a single-zone operation. the desired flow rate is scheduled with furnace
`input following the function shown in Figure 2.7. Flow rate decreases in proportion
`to input during the first half of tum-down. keeping temperature rise fixed. Then,
`during the last half. flow rate is constant allowing temperature rise to decrease.
`The break at the half-way point was chosen arbitrarily. A careful cost-benefit
`analysis would have to be performed to determine the best break point. To
`maximize energy savings. flow rate should decrease throughout the turn-down
`range rather than stopping half way. but adequate velocity pressure is needed to
`overcome duct resistance and to provide good distribution in the heated space.
`
`For multi-aone operation. the desired flow rate should consider the number
`of zones calling and their loads. Had this been done the control schedule would
`have been extended as shown by the dotted line in Figure 2.7. This would have
`allowed the system to operate at lower flow rates.
`
`2.2.4 Qombustlon Air Control
`
`A General Electric ECM was used to drive the inducer blower which was a
`Fasco "Swirlwind" unit.
`
`To maximize combustion efficiency. motor speed was reduced with firing
`rate. This was accomplished by reducing torque. A schedule of torque vs firing rate
`that was set up in the lab and then adjusted in the field always provided enough
`combustion air. As with the circulating air ECM. the inducer ECM also provided
`a tachometer output. but this output could not be used for closed—loop control
`because the fan characteristics of the Swirlwmd blower were not the same as those
`of the circulating air blower. A matched blower/motor set is needed for closed—loop
`control. The tachometer output. however. was used as a safety device as follows:
`At any given torque. a certain speed is expected.
`if at least that speed is not
`achieved. it can be assumed that the motor has failed and the gas should be shut
`off, This safety feature was used at the SMART HOUSE.
`
`2.2.5 Hard-Wired Control Logic
`
`The test furnace comes from the factory hardwired with control logic that
`properly interfaces the furnace with the thermostat and with safety devices.
`Figure 2.8 shows this in ladder-logic form. The thermostat. with its four terminals
`R. G. Y. and W. is inside the dotted box.
`
`15
`
`

`

`Fang
`
`.35.!mmno5anRmu
`
`mm
`
`6505ms:
`
`v;
`
`né
`
`«4.
`
`».w
`
`ad
`
`ad
`
`he
`
`9:
`
`ad
`
`«.0
`
`n6
`
`NA.
`
`.3
`
`
`
`$20:5:252:3vote“..835m9.35£3.22".:<95285.52up;3:30
`
`
`
`
`
`
`
`
`
`(spuesnou1)
`
`(was) Elva MO‘H NVJ
`
`16
`
`
`
`min0.5mm
`
`
`
`
`

`

`car
`
`MZON._._.4:S
`
`20F<mmmo
`
`MZONmSGz—m
`
`202.55%.
`
`
`38on8on88
`£323:".0.Fzmommause.532.
`
`SHE EauucumEtcou2mm30EL_<mama—autoWm959”.
`
`cor
`
`cm
`
`3
`
`2.
`
`on
`
`om
`
`3
`
`on
`
`(wnmwi :IO lNBOHad)31VH MO'H
`
`17
`
`

`

`TF84-291
`
`VOLTAGE
`
`(BURNER)
`
`HONEYWELL
`SGSGD
`
`CAUTION-
`HIGH
`
`Figure 2.8 Factory»Wired Control Logic
`
`18
`
`

`

`The sequence of operations in a normal heating cycle is as follows: Voltage
`is applied to terminal W when the thermostat calls for heat.
`lDR is energized
`through NPC and contacts 3~I and 6-4 close. When contacts 3-1 close. IBM is
`energized. When IDM has achieved a suilicient speed. NPC closes (contacts 6-4
`keep lDR energized while NPC closes). When NPC closes, the Honeywell ignition
`controller is energized provided LC. OTS. and EDS are closed. The ignition
`controller has a time delay for purging the combustion passages. After the purge
`period. the pilot is lit followed by the main burner. As the plenum heats, BFC
`closes. which energizes iBM.
`If HCR is energized. which happens when the fan
`on/ auto switch is in the on position and terminal G is powered. IBM will have been
`continuously energized and typically operating on high speed. The gas will stay on
`until either the thermostat is satisfied or one of the safety devices opens. As the
`
`plenum cools. BFC opens. which de-energizes IBM. This completes the cycle.
`
`The modulating furnace uses the same sequence of operations described
`above during startup and shutdown However. when modulating. NPC is locked out
`because inducer speed is reduced below its setpoint and would otherwise shut the
`gas off. Thus. NPC is only proved at startup.
`
`Tecogen‘s use of a microprocessor made it necessary to rewire the furnace.
`The new ladder-logic diagram is shown in Figure 2.9. As shown. the status of
`several switches is monitored by the microprocessor. For example. LC. OTS. and
`EDS are monitored via optically-coupled relays. Should these switches open. the
`microprocessor is programmed to halt operation and send an alarm signal to the
`operator.
`
`2.2.6 Single-Board Controller
`
`A microprocessor-based single-board controller was used at the SMART
`HOUSE for control and data acquisition. The control board is a standard Tecogen
`
`product that is used on its cogeneration and engine-driven chiller systems. The
`advantage of using a microprocessor-based system is the flexibility it offers for
`testing various control strategies.
`
`Tecogen's single-board controller combines the Motorola 68000 and
`Microware's UNIX-like. 059 operating system with quality hardware interfaces.
`089 is both a ROMable real time system and a full function disk-based system with
`
`a range of languages. The port includes complete high level software/hardware
`interfaces.
`‘
`
`This board in no way represents what is needed for a commercial unit but
`it offered the needed power during this development phase of the program.
`
`19
`
`

`

`
`
`
`
`10.25am29.50rm?—mum
`
`3”»th
`
`
`
`190853.3in
`
`
`
`>149»zen—.5300!c!
`
`52.5.5mm
`
`
`
`
`
`$9.5633.55cm>owho
`
`
`
`
`
`.552895¢mem>F<auzon:
`
`105SnExec-Eman
`
`
`
`mi;m<o:2:>6:
`
`
`
`u>4<>r40.54!>2
`
`40528.550.. ><._wc
` .55282E530393
`
`
`
`

`

`2.2.7 Data Acguiaition and Monitoring
`
`Data from the single-board controller were sent to an on-site PC. This
`allowed storage of the data for later analysis and provided for real-time viewing of
`the system‘s operation Several data collection and processing computer programs
`were written for the PC to complete data acquisition
`
`2.2.8 Zgne Control
`
`The zone control program determines furnace firing rate and controls the
`on/ofi‘ dampers for the supply and return air ducts for each zone. It tries to
`maintain zone temperatures at their respective setpoints under all conditions when
`a heating load exists. The algorithm has no provisions to handle a cooling load.
`Control "starts" when any zone's temperature drops below its setpoint. Control
`continues so long as the temperature doesn't exceed the setpoint by more than
`two degrees. Above two degrees. control "stops." The basic controlling equation
`used at the SMART HOUSE was:
`
`of... - 3 (we <5.) [19. E. + firm, at + 0...]
`
`.
`
`5
`
`l
`
`where:
`
`where :
`
`qum = desired furnace firing rate (percent of capacity):
`ZUJ
`= zone units of zone 1 (the estimated proportion of the total load due
`to zone 1);
`‘
`= 0 if Tame >
`“We“ + 2). (damper closed). or 1 if Tmne < Twink“.
`(damper open):
`
`81
`
`Ej
`Kp
`K1
`t
`cht
`
`= zone error (Tutpomt — Tum):
`= proportional gain constant:
`r:
`integral gain constant;
`= accumulated time after 81 becomes unity;
`= estimated heat load. or feed forward signal.
`
`The proportional gain is determined as:
`
`K9 3 QM mJEm-z
`
`= maximum furnace capacity (100%);
`qum max ll
`maximum error allowed (10°F).
`
`21
`
`

`

`
`
`The integral gain for a critically damped system is:
`
`K‘ =Kp/1:
`
`where:
`
`r = system time constant (500 seconds. as determined from experiment).
`
`The estimated heat load (9:39 is determined by two difl‘erent methods: UA
`and historical load. Tests were conducted on both methods.
`
`The equation for the UA method is:
`
`0. = UA 5w - rrm + sum» capacity)
`
`estimated overall heat transfer coeflicient it outside surface
`area:
`
`average value of zone setpoints:
`= outdoor air temperature adjusted to account for thermal
`capacitance of house;
`
`It
`
`maximum output of furnace (90 MBtu/ hr).
`
`where:
`
`(qum)i
`At
`
`[I
`H
`
`furnace firing rate at time Increment 1;
`sampling period;
`
`time period over which Qest is determined.
`
`22
`
`

`

`The value ZUj was selected for each zone to reflect the heat load of that zone
`relative to the total. When zone j calls for heat with other zones OK. it receives the
`heat output to match its load.
`
`It was assumed that zone