Home

PIER/NBI Report on FDD - Western HVAC Performance Alliance

1. Qualitative voc Monitoring Real Time Data Maintenance Repairs Service Performance Management Capital Equipment Replacement Predictive Planned Quantitative Performance A 10 iC3 Sensors Niagara AX FDSI Insight Data Analytics Niagara Ax Remote HVAC Remote HVAC FDSI Insight Industry Expertise FDSI Inside Niagara AX FDSI Insight System Tuning Niagara AX Remote HVAC Bi e Remote HVAC FDSI Insight Industry Expertise Real Time Data Predictive Planned Quantitative Performance Fault Diagnostics Optimized Minimized Remote Real Time Quantitative Intelligence Hovv do you or vvould you use FDD 1 10 11 12 13 14 Are you or your organization currently using any kinds of tools like FDD Can you imagine using FDD tools in your service business How What s the business model How would you change your service offering Would you be more interested in hard faults degradation faults or fault prediction Would you be more interested in monitoring fault detection or diagnostics What value would it bring to you reduce the number of trips required reduce the length of a service call What value would it bring to your customer amount of energy savings improved uptimer How much of a premium would your customers be willing to pay or how much of a discount would
2. 5 2 13 4 2 Definition of FDD 5 oas eeaeee naer 14 4 3 Description of Available and Developing Tools 16 4 4 Description of Potential FDD Approaches 18 Evaludilon of Appuaaslhes 34 5 1 Magnitude of Energy 3333 s s essesi iss s s s s 34 5 2 A EAT O conesececssccssecssecetecesseeeseesacenenenetenssonsecneesnecesaesuaecereaeseessasens 34 5 3 ES AAC O 35 5 4 Probability that Fault will Get Fixed ooooocconnnnccccnnnoncccnononnnonononococononcno nono nno con nn cocoa nn nccnann cnc 35 5 52 Reliability RODUSTA ot 35 5 6 Ease of Dep Medio A A A a 35 5 7 Other Maintenance 85 35 35 6 1 Highest Priority Approaches 36 6 2 NOXtStODS tt A A A A A ate 37 AA s n AR 37 1 Introduction In this second Phase of this project Fault Detection And Diagnostics Moving The Market And Informing Standards In California
3. Low e Reducing the time required for a service call A diagnostic tool that helped a service technician to do his or her job on the roof will have a benefit in reducing the time for a particular job This is not as great a benefit as avoiding a service call but if it allows for more jobs per day it is a benefit 4 Definition of FDD Approaches VVith a good understanding novv of the criteria for selecting appropriate FDD tools vve novv turn our focus to potential FDD solutions adapted to rooftop packaged air conditioning systems and heat pumps from 5 to 50 tons All of the listed solutions may not be available on the market the goal is to give the broadest possible view of the different approaches The objective of an FDD system is to detect faults early diagnosis their causes and enable correction of the faults before additional damage to the system or loss of service occurs This is accomplished by continuously monitoring the operating conditions and comparing them to a model of expected performance When actual measured operation does not match expectations determined with use of a model a fault is detected Ultimately the obiective is to ensure that the building ovvner or operator vvill respond in an appropriate manner remedying the problem VVe vvill define a method of classifying different approaches to FDD and evaluate their potential for use in the market or in Title 24 Building Standards 4 1 Classification of FDD Tools
4. O O O O o o o B 17 4 4 Description of Potential FDD Approaches For each cell in Figure 1 vve describe the follovving aspects of the hypothetical available and developing tools e Required information what information set does the system need to access Some of this information is readily available from standard higher end RTU control boards whereas other information is obtained by the mean of additional sensors e Example of detected faults this list of faults is based on publications detailing the operation of this FDD solution if available or engineering knowledge This is by no means exhaustive but serves as an illustration of the potentialities of the system e Hardware requirements based on the required information and the type of processing envisioned it is possible to list the necessary components and an approximate price if available 4 4 1 Refrigerant Quantitative Model Approach This approach works by comparing performance indices calculated by an internal model on one side and obtained by processing input values on the other side By observing the residuals it is possible to detect faults in the system Required information e Suction pressure can be substituted e Liquid pressure with additional Discharge Pressure temperature sensors e Suction temperature e Liquid temperature e Ambient temperature e Return air temperature e Supply air t
5. Public Interest Energy Research PIER Program PROJECT REPORT Fault Detection and Diagnostics Moving the Market and Informing Standards in California DRAFT E A Report of the Western Cooling Efficiency Center EFFICIENCY CENTER RESEARCH e INNOVATION PARTNERSHIP Prepared for California Energy Commission MARCH 2011 Prepared by Western Cooling Efficiency Center CEC 500 08 049 University of California Davis 1450 Drew Ave Suite 100 Davis CA 95618 wcec ucdavis edu Prepared by Primary Author s Kristin Heinemeier Mark Cherniack Julien Bec U C Davis Western Cooling Efficiency Center 1450 Drew Ave 100 Davis CA 95618 530 752 0280 wcec ucdavis edu CEC Contract Number 500 08 049 Prepared for New Buildings Institute and California Energy Commission Cathy Higgins NBI Project Manager David Weightman CEC Contract Manager Chris Scruton Program Area Technical Lead Virginia Lew Office Manager Energy Efficiency Research Office Laurie Ten Hope Deputy Director Research and Development Division Melissa Jones Executive Director the information in this report DISCLAIMER This report was prepared as the result of work sponsored by the California Energy Commission lt does not necessarily represent the views of the Energy Commission its employees or the State of California The Energy Commission the State of California its employees contractors and subcontractors
6. R Tenant aggravation maintenance is often a loss leader where contractors do bad installs and make up on service It s typical that a maintenance tech will get 20 hours week to do service and has to get as many customers as he can What problems are the most common or the most important why causes of service calls D Non PM filter low air freezups heat exchanger failure Depends on the quality of maintenance Can t bid 1 hour for preventive maintenance Motivation Try to avoid service call between now and next regular service See study by Jones Long LaSalle BOMA Studies R Electrical contactors expansion valve takes out compressor refrigerant leaks operating pressures Training issue don t overcharge D Economizers it costs 100 to fix the actuator The customer is still comfortable so it s a hard sell We use Honeywell software We talked with Adrienne Thomle the 7650 costs 20 and the 7660 costs 24 Have you ever told an owner that their RTU was on its last legs What symptom were you responding to What did the customer say What data did you need to make this call and convince the owner How long is a typical service call how many service calls can a technician make in a day What influences this A 17 9 R budget for capital improvement 20 energy savings Good relationship fan was looking wornout and there was bearing noise Checked on parts availability reduced disruption for tenants Othe
7. of codes is to get building owners to implement measures that the market for some reason or another is not successfully providing This would suggest that even if a product does not meet marketability criteria it can still be effectively implemented in Title 24 so long as it meets the criteria for code inclusion Ideally we would provide actual numerical metrics for these criteria lt should be noted however that most of the criteria are quite difficult to assess particularly when we are discussing hypothetical tools Diagnostic tools in general are difficult to assess because it is never known ahead of time how many and what type of faults will be detected It is beyond the scope of this project to provide quantifiable metrics for each criterion It should be possible however to compare different tools and describe characteristics somewhat qualitatively When we evaluate the different approaches we use a scale of High Moderate or Low to describe whether the criteria are met Fully Partially or Not at all 3 1 Magnitude of Energy Savings The energy savings expected from an FDD approach depends entirely on the type and number of faults that can be detected and addressed Reductions in greenhouse gas emissions go along with these energy savings Some of the faults that RTUs experience are indicated below along with an indication of the relative magnitude of energy savings possible through remedying the faults High e Compressor
8. they need to see to be provided this service One time capital cost monthly quarterly annual fee based How much would you be willing to pay to have this tool available What types of customers do you believe would be most interested in this kind of service Would you be interested in remote monitoring of RTUs dependent on customer size or would you like for all customers What performance monitoring FDD information do you think are most important in terms of optimizing performance efficiency What information would you like to see on an FDD PM GUl dashboard If you could design an RTU performance monitoring system with remote access what would be included as features Do you have any customers who you think we should interview A 11 Attachment 2 intervievv 1 Participants Erik Emblem E Cal SMACNA and Rick VVylie VV Beutler Corp December 18 2009 Background E A Consultant vvith the California Sheet Metal VVorkers Contractor background 7 years service 7096 of the client s activity Clients expectations Reliability dependability consistency Union 625 contractors in California 50 union members From fabrication companies to HVAC global suppliers VV Installation technician Done some service work Maintenance intended as maintenance agreement Commercial more likely to see the value of maintenance Some have their own maintenance staff filter change simple parts replacement etc VVh
9. 500 1000 to fix it If disabled everybody gets what he wants Possible bc cheap energy Q snap disk thermostat damper motor question of Mark regarding the link btw snap disks and motors not really understood VV whatever the piece economizer working is quite complex Setting it correctly is a challenge Q does the customer asks to fix the economizer VV not noticed by the customer Noticed during maintenance or customer complaining about low cooling Service technician fresh air over dramatized gt need for less fresh air ie less bills most of the time Customers often prefer to have it disabled than fixing it E ifthe customer knew the cost of disabling the economizer it could be possible to make a value proposition Right now they don t see the benefit W Long payback E lack of public awareness Have to do the check hard to convince W on the customer side people would rather spend on a solar system even if less efficient importance on metering the production Car analogy mpg information very educative Importance of metering Monetary conversion process EER to Hard to do with energy efficiency Have you ever told an owner that their RTU was on its last legs What symptom were you responding to What did the customer say What data did you need to make this call and convince the owner W 1 old and broken down for example compressor out Age and cost of repair taken into account n a house 12 years and
10. Deliverables B FDD Prioritization we identify and prioritize the faults that can be detected by a set of currently or shortly available diagnostic tools and evaluate the available tools In the next Phase we will develop a draft specification for new requirements for FDD in Rooftop Units We will also hold an industry roundtable to present the draft to a set of industry actors and obtain their feedback The following phase will consist of drafting a proposal for a standard and following the CEC Title 24 review process Background Remote and automated Fault Detection and Diagnostic FDD tools have the potential to save considerable energy in California s fleet of existing commercial rooftop air conditioning units RTUS The market for these systems has not yet materialized however Tools have been available for larger systems for some time although even these have not enjoyed a significant market share In RTUs there are fewer tools available and little to no market share Since RTUs cool over 70 of the commercial square footage in California they are a significant source of energy consumption and peak demand Under the best of circumstances RTUS are not as efficient as larger built up systems However in reality they are even less efficient Many market failures have led to a lack of quality in installation and maintenance of these units and their performance is suffering Most RTUs have some sort of fault that is increasing t
11. Outdoor Air Flow SETPOINT or check BAS Network interface card status and wiring Supply Air Temperature The unit is reading a signal These unit functions are PAR An automatic reset Sensor Failure that is out of range for the disabled occurs after the designated Problem The Supply Air Supply Air Temperature input a Supply AirTempering Supply Air Temperature input Temperature sensor input is lon the RTM Temperature lt b Economizing returns to its allowable range out of range 55 F or Temperature gt 209 F c On CV units the Supply Air In order to prevent rapid Temperature low limit function cycling of the Diagnostic there k i Sansor resistance is disabled is a 10 second delay before should be between 830 ohms 200 F and 345 7Kohms 40 d On VAV units the Supply the automatic reset F Check field unit wiring Air Temperature Control between Sensor and RTM heating and cooling functions are disabled Return Air Temperature The unit is reading a signal The Economizer Enable r e PAR An automatic reset Sensor Failure that is out of range for the Enthalpy function reverts to occurs after the RA Problem On units with the Return Air humidity sensor Reference Enthalpy Temperature input returns to Comparative Enthalpy option Temperature 55 F or changeover Level 2 its allowable range the Return Air Temperature Temperature gt 209 F control continuously for 10 seconds sensor input i
12. a dirty filter None 59 Gas valve 1 not energized 3 default times 2 minutes Only action taken is storing code in memory after a demand Check gas supply ignition control and wiring ECTO 3 09 GV1 75 Outdoor Temperature RT17 Sensor Problem Check The control defaults to a high outdoor temp operation wiring and sensor 91 Outdoor enthalpy sensor A7 open Check sensor and No economizer free cooling operation if economizer wiring mode is set to ODE or DIF 92 Indoor enthalpy sensor A62 open Check sensor and No economizer free cooling operation if economizer wiring mode is set to DIF 93 The control has changed the system mode because of IMC has switched over to the backup mode option set an error with the controlling sensor or because of a with ECTO 6 01 loss of communication 99 Outdoor Air Control Sensor A24 open Cleared by No OAC operation Damper closed to minimum IMC reset position York Alarm Codes Alarm Problem Alarm 13 indicates the Space Temperature Sensor has failed open or shorted Alarm 14 indicates the Supply Air Temperature Sensor has failed open or shorted Alarm 15 indicates the Return Air Temperature Sensor has failed open or shorted Alarm 16 indicates the Outside Air Temperature Sensor has failed open or shorted Alarm 17 indicates the Dirty Filter Switch has tripped Alarm 1A indicates a microelectronics failure and the control is o
13. about energy consumption If you could design an RTU performance monitoring system with remote access what would be included as features E VV Different information levels idiot light alarm if efficiency goes under a certain threshold efficiency for each unit log all activity 14 Do you have any customers who you think we should interview E see about national restaurant association A 16 Attachment 3 intervievv 2 Participants Denny Mann D Marina Mechanical and Russ Donicci R Mechanical Air Service December 21 2009 Background D Service S4Million and Construction R 2Million business 80 construction Residential component is high end Service retrofit clean rooms 60 new 40 service What is your service business like 1 What constitutes RTU service in the field D Maintenance contracts recurring revenue business year round General operations for retrofit Commercial light industrial Not much refrigeration 14 trucks Total cost not hourly rate Contract lt 100 000 sqft Unless on staff 95 on contract R It s hard to get a maintenance contract How is your service business structured D referrals common Contract for 4 times per year 500 pre approved avoids truck roll Full coverage guarantee 3 times the price of maintenance Q Callbacks different between manufacturers D No Tho a few years ago York had gt 20 D we provide a Warranty of 1 year
14. above compelling to replace it Another significant cause could be a hard to fix coil failure Window of opportunity with fluid change gt if some equipment is not on sale anymore need to change the whole unit More significant with split system because each half is sold independently If one needs to be replaced and only the new refrigerant type is available then the other needs replacement too Other case many units in a complex for example Can be planned during the year Dismounted equipment used as spare for the rest of the units On RTU parts available for a long time Fluid available Concern for change if many RTUs at the same location and some are dead All may be replaced to avoid a mix of refrigerants 2 part of a building renovation Reroof ownership change tenant change Financial transaction already taking place Part of the negotiation to extend or renew the lease Depreciation calculation by large property owner Q how to assess the remaining life on the unit Customers arrive with preconditioned ideas Like under 500 policy Q consider energy operating costs W not directly fixed changed if broken gt likely to go with the fastest repair Energy efficiency is a secondary argument E Carfax like report click a few things and gives you a number A 13 10 11 VV best customer maintenance customer Long term view Analysis on a whole building level Q what arguments to get people enroll i
15. air or mixed air sensor failure Faulty repairs Low changeover temperature setpoint Use of a single stage cooling thermostat The energy impact can be estimated between 14 and 40 according to whether the economizer is malfunctioning or not functioning at all 2 2 3 Thermostat errors e Improper thermostat single stage cooling only e Cycling fans during occupied periods B 5 e Continuous fans during unoccupied periods e Improperly installed resistors e No nighttime setup or setback Savings can vary Correcting cycling fans during occupied periods will even increase the energy use but improve indoor air quality The highest savings up to 40 will occur when the thermostat is preventing the economizer from operating 2 2 4 Failed sensor This problem has been noted in 20 of the units The energy savings for repairing sensors vary over a wide range They can be modest if the value is slightly incorrect but can go up to 40 if they enable a non functioning economizer 3 Criteria for Evaluating FDD Approaches In order for an FDD approach to be viable in the market and to be justifiably included in Title 24 it must meet several criteria The primary criteria for Title 24 are energy savings and cost effectiveness The primary criteria for the market are somewhat different however To be successfully marketed the tool must be low cost marketable and reliably detect important problems in buildings Of course the role
16. and that the equipment make and model includes factory installed hardware that match the information specified on the manufacturers cut sheets and design plans The functionality of the FDD must also be tested in the field Form MECH XXA is used to verify that the criteria are met Construction Inspection 1 The following sensors should be permanently installed to monitor system operation and the controller should have the capability of displaying the value of each parameter Refrigerant Refrigerant Air Relative Air Pressure Temperature Humidity Temperature e Suction Line e Suction Line e Outside Air e Outside Air e Liquid Line e Liquid Line e Supply Air e Return Air e Supply Air The controller will provide system status by indicating the following conditions Compressor Enabled Free Cooling Available e Heating Enabled e Economizer Enabled e Mixed Air Low Limit Cycle Active The unit controller shall have the capability to manually initiate each operating mode so that the operation of compressors economizers fans and heating system can be independently tested and verified The unit controller shall have the capability to detect at least ten of the following faults Air Temp Sensor High Refrigerant Charge Low Refrigerant Charge Failure Fault Compressor short cycling Refrigerant Line Restrictions Refrigerant Line Non TXV Problems Condensables Lovv Side HX problem High Side HX problem Capacity D
17. be substituted e Liquid pressure with additional e Discharge Pressure temperature sensors e Suction temperature e Liquid temperature e Ambient temperature e Return air temperature e Supply air temperature Example of detected faults e Capacity to detect the following faults e Refrigerant leakage e Compressor valve leakage e LLrestriction e Condenser Fouling e Inadequate airflow Hardware requirements amp Pricing elements e Necessary addition of pressure sensors and or additional temperature sensors gt material and installation costs e Data logger Simple signal processing hardware for timeseries based approach e Communication module required for any FDD 4 4 4 Air Quantitative Model Approach By simulating the behavior of the unit for the measured input conditions and comparing the calculated and measured supply air temperature it is possible to detect faults vvithin the system Required information outdoor air temp OAT return air temp RAT mixed air temp MAT supply air temp SAT fan status cooling heating mode operation outdoor air damper position Example of detected faults supply air temperature too high inadequate airflow supply air temperature too low incorrect refrigerant charge compressor leakage Hardware requirements amp Pricing elements temperature sensors data processing module in order to run the physical model Communication module required for any FDD 4 4 5
18. e Energy Related Environmental Research e Energy Systems Integration e Environmentally Preferred Advanced Generation lIndustrial Agricultural VVater End Use Energy Efficiency e Renewable Energy Technologies e Transportation Fault Detection and Diagnostics Moving the Market and Informing Standards in California by the Western Cooling Efficiency Center is an interim deliverable for the Fault Detection and Diagnostics Moving the Market and Informing Standards in California Program conducted by New Buildings Institute contract number 500 08 049 The information from this project contributes to PIER s Buildings End Use Energy Efficiency Program For more information about the PIER Program please visit the Energy Commission s website at www energy ca gov research or contact the Energy Commission at 916 654 4878 ABSTRACT This document consists of three distinct reports One crucial part of the prioritization of FDD tools is collecting intelligence from key stakeholders In the first section of this report A Interviews with Key Stakeholders we describe the process of developing an intervievv guide and carrying out a small set of intervievvs VVe summarize the intervievvs that vvere held as vvell as provide the detailed responses to our list of questions n the second section of this report B FDD Prioritization vve identify nine different potential approaches depending on the type of data collected air side refrigerant sid
19. effective high pressure refrigerant can either e Leak back into the suction line across the suction valve or e Leak back into cylinder across the discharge valve This corresponds to a loss of volumetric efficiency the impacts of which on efficiency and COP have been evaluated by in Breuker and Braun 1998 2 1 2 Condenser coil fouling This fault is caused by a buildup of debris on the condenser coil This limits the available condenser coil area and reduces heat transfer and the total airflow across the coil 2 1 3 Inadequate Airflow This is caused by a buildup of debris on the evaporator coil or other restrictions in the airpath The consequences are twofold the airflow across the evaporator is reduced and the heat exchange efficiency is limited The impact of the second effect is limited though B 4 Airflow has been found to be deficient in 42 of the cases The following criterion has been used the airflow is considered too low under 300CFM ton to be compared with the 400CFM ton used for industry efficiency ratings 2 1 4 Liquid line restriction This fault occurs when a filter dryer or expansion device is obstructed by debris This increases the total pressure loss in the liquid line 2 1 5 Refrigerant leakage undercharge overcharge This fault can either be caused by a slow leak in the system or by the wrong amount of refrigerant introduced into the system Charge assessment is not straightforvvard and the different meth
20. for California Energy Commission MARCH 2011 Prepared by Western Cooling Efficiency Center CEC 500 08 049 University of California Davis 1450 Drew Ave Suite 100 Davis CA 95618 wcec ucdavis edu DEVELOPMENT OF A DRAFT FDD STANDARD FOR ROOFTOP UNITS In this third Phase we have developed a draft specification for new requirements for FDD in Rooftop Units We have also held an industry roundtable to present the draft to a set of industry actors and obtain their feedback The next phase of activity described in a separate report will consist of drafting a formal proposal for a standard and following the CEC Title 24 review process Remote and automated Fault Detection and Diagnostic FDD tools have the potential to save considerable energy in California fleet of existing commercial rooftop air conditioning units RTUs The market for these systems has not yet materialized however Tools have been available for larger systems for some time although even these have not enjoyed a significant market share In RTUs there are fewer tools available and little to no market share Since RTUs cool over 70 of the commercial square footage in California they are a significant source of energy consumption and peak demand Under the best of circumstances RTUs are not as efficient as larger built up systems However in reality they are even less efficient Many market failures have led to a lack of quality in installation and maintenanc
21. for RTUs The guide is provided in Attachment 1 The guide consists of three separate sections Section 1 What is your Service Business like This is intended to find out about typical service models and the business environment into which FDD will have to fit No mention is made in this section of FDD e Section 2 Here s FDD This is intended to give the interview participants a common grounding in what is meant by FDD tools e Section 3 How Do or Would You Use FDD This section is intended to establish whether or not the participant has used any FDD tools and why or why not It is also intended to identify the necessary characteristics that would make FDD usable by them A 4 Summary of Interviews Tvvo intervievvs vvere held on December 18 and December 21 2009 respectively Each of these intervievvs included tvvo participants Attachments 2 and 3 present the ravv responses from the intervievv participants Some of the major themes that emerged from these interviews include What is your service business like e How important is the service contract Os Oy 0 o Best business is through repeated relationship Maintenance contracts offer recurring revenue business year round Referrals common Some firms offer a full coverage guarantee 73 times the price of maintenance Retention rate is important It s hard to sell a maintenance contract Some firms offer a 2 year warranty on insta
22. it does not require remote access FDD tools used in commissioning typically do not require historical data although Performance Monitoring data can be used to supplement the commissioning tool It is difficult to envision how commissioning tools would be included as a requirement in Title 24 unless commissioning interventions were also required B 10 There are already tools available that assist service providers in diagnosing a system in the context of a commissioning like process The Honeywell Service Assistant provided to the market by various California utilities through their Air Care Plus programs is one example lt is difficult to imagine the building code requiring an intervention such as this although it could require enabling technology that would make these kinds of interventions possible They would not generate savings on their own hovvever Maintenance Avoid failures or degraded operation and optimize running costs by optimizing the maintenance operations The match betvveen these situations and the available FDD methods described later in section 4 can be summed up as follovvs in Table 3 Table 3 Assessment of Deployment Capability for Different Types of Models Deployment Method Quantitative Qualitative Timeseries Performance monitoring High Moderate Moderate Commissioning High Moderate Low Maintenance High High High Quantitative approaches seem to give the best performance but
23. model and can therefore be used on any kind of refrigerant system It allows the user to assess system performance based on temperature pressure and power measurements Because of its cost this tool is applied most frequently as a service productivity tool rather than an ongoing degradation detector It is typically used in a service offering in which an experienced technician uses the tool to efficiently assess the condition of one or more RTUs at a facility An evaluation of possible savings documented in Energy Optimisation Potential through Improved Onsite Analysing Methods in Refrigeration Arul Mike Prakash 2006 found that out of 49 analyzed air conditioning systems 44 were found with faults The most frequent faults detected were related to charge 70 of units expansion valves 30 and secondary flow 11 No values on efficiency are documented but an average variation in COP of 11 5 was documented Cost A complete unit is around 5000 S6000 Enerey savings Case study An evaluation of possible savings is documented in Energy Optimisation Potential through Improved Onsite Analysing Methods in Refrigeration Arul Mike Prakash 2006 The most frequent faults vvere related to Charge 7096 of units expansion valves 3096 and secondary flow 11 Out of 49 analyzed air conditioning systems 44 have been found with faults No values on efficiency are documented but an average variation in COP of 11 5 has been doc
24. participants both in person and those remotely was wide and deep limiting this report to a high level summary What follows are key takeaways and issues identified to follow up on for both the PIER and CASE team researchers We will be contacting individual participants for further discussion about comments made at the Roundtable and those received pre and post roundtable We very much appreciated the positive tone of the meeting This is a complex topic as all of you all understand There are different needs for different markets that have to be accounted for and balanced in setting standards For example Walmart needs an approach to manage the 25 million HVAC alarms generated company wide in 2009 The company needs intelligence that manages and reduces alarms not necessarily additional sensors and algorithms that add to them Compare this with the owner of a suburban 10 000sf two story office building with a handful of RTUs who would benefit from basic FDD functionality in the RTUs such as monitoring airflow economizer operation sensor malfunction refrigeration charge status and not a whole lot more The tenor of the meeting was very positive Overall the researchers got the message to e Proceed with the potential for a 2011 Title 24 Prescriptive Measure e Provide substantiation of o FDD energy savings benefits and persistence o FDD product availability in the market or on the way o FDD product costs including communications gateway o Fa
25. roof for remote viewing Webenabled is a real option Title 24 standard should not be too prescriptive Try to specify performance objectives within the prescriptive measure This can drive new technology solutions FDD should be integrated with Quality Installation practice C 5 Quotable Quotes q The California Reach Code is not a dumping ground for unsupported measures r FDD needs to be developed into an iPhone or iPad app s Innovation comes into the market at the speed of profits Next Steps The next steps in this research project to be described in future documents include Quantify the expected energy savings and costs of systems By conducting a survey and analysis of system costs and simulating the savings attributable to FDD we can estimate the cost effectiveness in a typical building This modeling may also be used to determine savings from implementing FDD in a specific RTU for compliance purposes This work will be done in parallel with the next step and will be complete in the first quarter of 2011 Discuss with industry stakeholders to get additional feedback on the draft requirements The draft standard will be disseminated widely and feedback from a range of stakeholders will be taken into account in developing a Code Change Proposal This will be done in parallel with the first step and will be complete by the end of the first quarter of 2011 Develop a formal Code Proposal Once all possible i
26. see 2 VV nterventions may be more efficient The system must actually bring vvhat it promises 6 What value would it bring to your customer amount of energy savings improved uptime E better interaction with the contractor Shows professionalism W long term performance monitoring may help to prepare the client for heavier preemptive interventions 7 How much of a premium would your customers be willing to pay or how much of a discount would they need to see to be provided this service One time capital cost monthly quarterly annual fee based W 10 for a very good field diagnostic device 8 How much would you be willing to pay to have this tool available VV to many possible situations to give a figure 9 What types of customers do you believe would be most interested in this kind of service A 15 10 11 12 13 VV higher end workers like doctors lawyers High end restaurants Retails with food safety issues Would you be interested in remote monitoring of RTUs dependent on customer size or would you like for all customers W only way to justify an added cost If technician has to show up even a good solution should cost less than 100 see 7 What performance monitoring FDD information do you think are most important in terms of optimizing performance efficiency What information would you like to see on an FDD PM GUI dashboard W An indication about performance another one
27. than optimal performance 8 What tools do the OEM s provide for diagnosing faulty operation 9 What tools do the OEM s provide for performance monitoring 10 Do your techs use these tools If not why not 11 What related to your RTU service business keeps you up at night A 8 Here s FDD 19349 21A19 0 9JOUIY JO je 3311195 3ULINP UBNIUYI HNP doul q Aew asnes Asea s jne4 uon dsu pue sulo3dul4 10 S 9pojA uon dsui pue s31139u adueWOjed 21151 12 jo pua ayy 1e3u S zey HUN e Ain pi UO DIP3Ad ne 3ululiN 2360 YM 2 0 9J0113y 10 E2 921435 3ulinp URNIUYIA 491492 9914195 0 00 asnpasoid dads 10 p N HNP 340W aq Aew sne Asea s13 ney soja Joswoydws oj sjapow PUR ou Hojwog SUE a duis pue siosuas Suloluoyy DIUPUIJOJ II payadxa se Jou 114 UO JIUN JO 550 ON yun ajoym uopuny e 10 5501 yney uoljepeisaq Atel sy ney uonenunuuy 52115048 1 uon2 q BULIOUUOA 00110112590 911039 8 Bank of America Intelligent Command and Control Center Gen 1 Gen 2 Remote Monitoring Maintenance Repairs Service Performance Management Capital Equipment Replacement Remote None Preventative
28. their use in the field seems unlikely due to the complexity and computational costs associated with them The other two types of approaches should be able to provide the required performance for the maintenance deployment but will provide very different results in term of commissioning and operation optimization Since the process timeseries method relies on previously logged data it will not be able to detect any fault initially present in the system This is quite a big limitation since much anecdotal evidence seems to show that errors are often introduced in the system during the installation of the units A qualitative model on the other hand does not rely on a baseline operating condition to evaluate the behavior of the system and could therefore diagnose errors at any time of the lifecycle of the unit The performance monitoring aspect though being linked to numerical values of efficiency criteria cannot be included in a qualitative model whereas a process timeseries based model can definitely detect a variation of performances due to an operation optimization If neither commissioning nor operation behavior evaluation want to be left aside some hybrid approach combining both methods could be envisioned 3 7 Other Maintenance Benefits While not valued in the cost benefit analysis done for Title 24 most FDD tools also provide maintenance benefits to the customer service contractor or both This is key to marketability of FDD T
29. valve leakage of course HVAC refrigerants currently in use are themselves notable sources of greenhouse gases and the detection of leaks when they occur can provide as significant if not more significant a reduction in greenhouse gasses e Condenser coil Fouling e Inadequate airflow e Economizer malfunction e Economizer misapplied e Other controls problems Moderate e Refrigerant leakage undercharge overcharge e Liquid line restriction Low e Short cycling 3 2 Cost Effectiveness The cost effectiveness of a diagnostic approach is dependent upon the potential savings of course But it is also dependent upon the cost to implement the method The cost is based upon the type of data that is required the overall number of points required any processing capabilities that must be added and communications hardware and access The principal cost incurred for FDD is for data collection Depending on the method that is used existing sensors installed in the RTU might be used Care must be taken to ensure that the sensors are of sufficient accuracy and are installed in the appropriate location In some cases redundant sensors might be needed to take the place of the existing sensors The cost effectiveness of some of the sensors that are typically used are High e Fan On Off e Compressor On Off e Operating Mode Cooling Heating Ventilation e Outdoor Air Damper Position Current kVV Moderate e Suction temperatur
30. 11 Prepared by Western Cooling Efficiency Center CEC 500 08 049 University of California Davis 1450 Drevv Ave Suite 100 Davis CA 95618 wcec ucdavis edu 1 2 7 CONTENTS A AAA O 3 86 3 2 1 Mechanical FUESE 4 2 2 lt Controls Faults sci n adh nik AS ea A ae A a a 5 Criteria for Evaluating FDD 6 3 1 Magnitude of Energy sss s asas s s s esi 333333 6 3 2 Cost Effectiveness cancer 7 3 3 Frequency of Fault A aa enmarcan EMMIS cccsceseesenstenecencersessnsnececeucedeesenaes 8 3 4 Probability that Fault will Get 55 5 s s 9 3 5 Reliability Robustness 2 2222 10 3 6 Ease of 55 55 149 2 10 3 7 Other Maintenance 5 52 11 Definition of FDD Approaches s s ee si ee ii si ii sis iyisi iris iii sis 12 4 1 Classification
31. Air Qualitative Model Approach This approach vvould get rid of the physical model of the quantitative approach and rely on the variation of parameters and predefined acceptable ranges for each parameter to detect a faulty operation This vvould be much simpler but vvould only be practicable during steady state operation Required information e outdoor air temp OAT e return air temp RAT e mixed air temp MAT e supply air temp SAT e fan status cooling heating mode operation e outdoor air damper position Example of detected faults e incorrect economizer set point e supply air temperature too high inadequate airflow e supply air temperature too low incorrect refrigerant charge compressor leakage Hardware requirements amp Pricing elements e Temperature sensor e Data processing module e Communication module required for any FDD Example Available Tool Smart Monitoring and Diagnostics System Description Battelle Pacific Northwest Division in collaboration with NorthWrite Inc has developed a tool for continuously monitoring the condition and performance of packaged air conditioners and heat pumps The Smart Monitoring and Diagnostic System SMDS is mounted in a small box installed on the side of each packaged air conditioner or heat pump and provides continuous remote monitoring and diagnostics for the unit It requires the following components e Temperature sensor e Data processing module e Communicati
32. C UCM S REACTION RESET REQUIRED Low Refrigerant Charge Low refrigerant Charge Ckt1 A Lockout Ckt1 and or PMR A manual reset is Ckt1 and or Ckt2 and or Ckt2 The cooling Ckt2 request is issued to the required anytime after the Problem The difference circuit has been active for 10 compressor Staging Control Diagnostic is set The between entering and leaving Minutes and the difference in Function Diagnostic can be reset by the evaporator temperatures is the entering and leaving Human Interface or Tracer or greater then Evaporator evaporator temperatures is by cycling power to the RTM Temperature Differential greater than the Evaporator SETPOINT for 10 minutes Temperature Differential Check Status of refrigerant SETPOINT for 10 continuous charge and running pressures minutes on affected circuit Minimum Outdoor Air Flow Minimum Outdoor Air Flow The active Minimum OA Flow PAR An automatic reset SETPOINT Failure SETPOINT out of range SETPOINT reverts to the occurs after the Minimum OA flow is lt 0 or gt Max Unit default Minimum OA Flovv Flow SETPOINT input returns Problem The input designated Airflow 46000 CFM for 90 and SETPOINT to within range for 10 as the SETPOINT is out of 105 Ton and 58500 CFM for continuous seconds or after a 120 Ton and up different valid Minimum OA range Flow SETPOINT selection ae source is user defined Check Wiring to the GBAS input designated for Minimum
33. Change of mean Anomalous transient 1 requires refrigerant flow or compressor map and head pressure 2 requires air side flow and RH Hardware requirements amp Pricing elements e Single phase diagnostic can be implemented in a low cost 150 MHz PC on a chip with a two channel A D converter with an incremental RTU manufacturing cost as low as 200 e Communication module required for any FDD Example Available Tool Virtjoule Description This system would be composed of RTU power sensors communicating with the Virtjoule gateway using ZigBee protocol This allows the information to be shared with Virtjoule network operation center NOC Data is then processed and communicated to the end user via different possible channels web service email SMS Approach The basis for the analysis is the usage metering and power signature provided by the power sensor This raw information routed to the NOC is then combined with additional data such as historic data or local weather information in order to detect and classify faults Cost Typical ROI lt 1 5yr Example Available Tool Non Intrusive Load Monitoring The Non Intrusive Load Monitoring NILM device developed by the Massachusetts Institute of Technology is an example of power monitoring tool that uses timeseries based methods to diagnose system condition This approach was first developed to investigate the energy use in whole buildings It has been dubbed no
34. Johnson Controls Don Langston Aire Rite Air Conditioning amp Refrigeration Inc Richard Lord Carrier Corporation Mike Lubliner Washington State Energy Extension Jim McClendon Walmart Jon McHugh McHugh Associates Jeff Miller California Energy Commission John Proctor Proctor Engineering Group Mark Rehley Northwest Energy Efficiency Alliance Todd Rossi Field Diagnostic Services Inc Chris Scruton California Energy Commission Vern Smith Architectural Energy Corporation Danny Tam California Energy Commission Stuart Tartaglia Pacific Gas amp Electric Buck Taylor Roltay Services Inc Adrienne Thomle Honeywell Matt Tyler Portland Energy Conservation Inc Anne Wagner Pacific Northwest National Lab Stuart Waterbury Architectural Energy Corporation Mike Walker Lennox Industries Larry Wei Lennox Industries David Weightman California Energy Commission David Yuill Purdue University Herrick Laboratory C 8
35. There are two ways of characterizing FDD tools one is based on the set of measured data points the tool uses to perform its analysis another is to look at what sort of model is used to determine expected performance Each of these factors has a strong influence on the final capabilities of the system 4 1 1 Classification Based on Input Data The input data acquired naturally falls into three distinct groups Airside Data Refrigerant Cycle Data and Power Data Air side e outdoor air dry bulb temperature return air dry bulb temperature e mixed air dry bulb temperature e outdoor air damper position signal supply fan status e heating cooling mode Outdoor air relative humidity for enthalpy based economizers e Return air relative humidity for differential enthalpy controls Refrigeration cycle e Suction pressure e Liquid pressure e Discharge Pressure e Suction temperature e Liquid temperature e Ambient temperature e Return air temperature e Supply air temperature Some methods substitute the pressure monitoring with additional temperature sensors Breuker and Braun 1998b Power Monitoring e Fan Compressor or RTU kW e kWh e Power factor Real and reactive power e Power quality 4 1 2 Classification Based on FDD Model The different approaches can also be characterized by the models used to determine expected performance For example for a first principle based or engineering model physical laws a
36. a draft specification for new requirements for FDD in Rooftop Units We will also hold an industry roundtable to present the draft to a set of industry actors and obtain their feedback The following phase will consist of drafting a proposal for a standard and following the CEC Title 24 review process Background Remote and automated Fault Detection and Diagnostic FDD tools have the potential to save considerable energy in California fleet of existing commercial rooftop air conditioning units RTUs The market for these systems has not yet materialized however Tools have been available for larger systems for some time although even these have not enjoyed a significant market share In RTUs there are fewer tools available and little to no market share Since RTUs cool over 70 of the commercial square footage in California they are a significant source of energy consumption and peak demand Under the best of circumstances RTUs are not as efficient as larger built up systems However in reality they are even less efficient Many market failures have led to a lack of quality in installation and maintenance of these units and their performance is suffering Most RTUs have some sort of fault that is increasing their energy use If these faults could be found and addressed then a significant energy savings could be realized Interview Guide The WCEC and NBI team developed an interview guide intended to gauge the acceptability of FDD tools
37. an to move the industry forward so that tools meeting these requirements can be developed and the requirements can be adopted in the next round of revisions 7 Bibliography Armstrong P R C R Laughman S B Leeb L K Norford 2006 Detection of Rooftop Cooling Unit Faults Based on Electrical Measurements HVAC amp R Research Vol 12 No 1 January Breuker M S and J E Braun 1998 Common Faults and Their Impacts for Rooftop Air Conditioners HVAC amp R Research Vol 4 No 3 July Cowan A 2004 Review of Recent Commercial Rooftop Unit Field Studies in the Pacific Northwest and California New Buildings Institute report to the Northwest and Conservation Council and Regional Technical Forum White Salmon WA Prakash Arul Mike 2006 Energy Optimisation Potential through Improved Onsite Analysing Methods in Refrigeration Master of Science Thesis Department of Energy Technology Royal Institute of Technology Stockholm Sweden Rossi T 2004 Unitary Air Conditioner Field Performance International Refrigeration and Air Conditioning Conference Purdue University West Lafayette Indiana July 12 15 Public Interest Energy Research PIER Program PROJECT REPORT Fault Detection and Diagnostics Moving the Market and Informing Standards in California C Draft FDD Standard for Rooftop Units A Report of the Western Cooling Efficiency Center WESTERN COOLING EFFICIENCY CENTER Prepared
38. are required This can be accomplished by tying into the building s Energy Management System or installing a dedicated modem and phone line It is often possible to use a gateway to allow the diagnostic module to piggy back on the building s communications infrastructure to reach the internet 3 3 Frequency of Fault A diagnostic tool is not as useful if it only detects faults that occur rarely The more common the fault is that it detects the more appropriate it is for Title 24 The faults that are frequent in RTUs are described in Table 1 and are provided below along with an assessment of their relative frequency High Inadequate airflow e Refrigerant leakage undercharge overcharge e Economizer malfunction e Economizer misapplied Moderate e Condenser coil fouling e Short cycling e Other controls problems B 8 e Locked rotor e Liquid line restriction e Liquid slugging Moderate e Loss of phase e Unbalanced voltage e Compressor valve leakage e Flooded start 3 4 Probability that Fault will Get Fixed It is easy to envision a code that requires manufacturers to install one of these types of FDD in their units in the factory or contractors to install after market tools that will identify faults However it is somewhat more difficult to envision how these tools can be guaranteed to provide savings After all they merely note the fact that a fault exists they do not fix it Some method must be found to
39. at is your service business like 1 What constitutes RTU service in the field VV on the phone ask for shovv up cost Some easy things can be fixed on the spot For the rest price proposition for fixing No need for permits for part svvaps 2 Hovv is your service business structured VV Business through repeated relationship E lot of contact through the internet Relationship established from quite some time family recommendations inter generational Sticker on the heater Nate study demographics on the first call gt more feminized VV New commercial practices price based competitive bid on maintenance contract If maintenance contract typically repairs not taken into account but discount applied Some higher end specific application may have a partial coverage of repairs 3 What problems are the most common or the most important why causes of service calls W less leaks than in split systems In order of occurrence Belt problems bearings contactors capacitors fan motor compressor rare but significant Missing filters Q how do you know bearings are shot VV from the noise can be seen More generally accurate service allows to see a lot Q what if it could be remotely detected VV possible to put sensors on everything but would have more issues with the sensors than with the parts in the first place A 12 Q VVhat about economizer issues VV first thing you disable Complexity vs benefits
40. ation calculation by large property owner Customers do not consider energy operating costs directly fixed changed if broken gt likely to go with the fastest repair Energy efficiency is a secondary argument What are the maintenance cost constraints o Can t bid 1 hour for preventive maintenance Motivation Try to avoid service call betvveen novv and next regular service It s typical that a maintenance tech will get 20 hours week to do service and has to get as many customers as he can typically 5 6 calls a day Compressor replacement 4 hrs but simple maintenance and diagnostics 4 6 day better diagnostics on the system makes you faster and more confident we try to do 2 1 2 calls day lt 10 tons 5 calls if they re simple if its superheat pressure or economizer 3 4 per day OEM s provide some things for diagnosing faulty operation Not much on the cooling side usually used by technicians if present but not very predominant right now How do you or would you use FDD Are you using tools now Most organizations currently not using any kinds of tools like FDD One vve intervievved vvorked vvith three tools in Air Care Plus They vvere hit and miss A lot of the reports didn t make sense t vvorks better vvith small equipment eg temperature probes Depends on ambient and return conditions A lot of cables and connections It s hard to be profitable Existing customers will ask why didn t you find i
41. business How What s the business model How vvould you change your service offering R some customers maintain different things Degradation re remote service assistant Change more level of service Clerk can t interpret data Customer predictive dovvntime energy costs VVe are trying to get into the energy component AB1103 requires the Energy Star Rating it s hard to get things like the number of PCs Carbon footprint is a motivator Need a 5 year SPT on equipment Would customers pay a 5 10 premium If they have a critical environment Retantion rate is important Need a more sophisticated customer Service assistant 3000 Tool needs to be portable Can t reduce service calls to 2 year because of filters and belts We offer a 2 year warranty on installation is customer signs a service contract A 18 10 11 12 13 14 VVould you be more interested in hard faults degradation faults or fault prediction VVould you be more interested in monitoring fault detection or diagnostics What value would it bring to you reduce the number of trips required reduce the length of a service call What value would it bring to your customer amount of energy savings improved uptime How much of a premium would your customers be willing to pay or how much of a discount would they need to see to be provided this service One time capital cost monthly quarterly annual fee based How much would you be wi
42. d Diagnostics Inc Sensus MI Available Air Qualitative University of Nebraska ClimaCheck Available Refrigerant Quantitative ClimaCheck Inc SMDS Pilot Air Qualitative Pacific Northwest National Lab NILM Pilot Power Qualitative Massachusetts Institute of Technology Low Cost NILM Pilot Power Timeseries Massachusetts Institute of Technology Sentinel Insight Beta Refrigerant Quantitative Field Diagnostics Inc Virtjoule Developing Power Timeseries Virtjoule Inc Low Cost SMDS Developing Air Povver Timeseries Pacific Northvvest National Lab 3 Draft FDD Standard for Rooftop Units n order to develop a draft specification for a RTU FDD standard vve have undertaken several tasks VVe revievved the current optional requirements for FDD in Title 24 2008 to determine the level of specificity that the new requirements might entail We developed a short summary of the existing standard that includes all of the requirements contained in the Standard the ACM the User s Manual and the Acceptance Test Forms We revised this summary to include the new requirements we felt might be included in a new standard The document shown in Attachment 1 is this summary NOTE that this Proposal was updated Feb 2011 after subsequent stakeholder input Industry Roundtable An Industry Roundtable was held at the Western Cooling Efficiency Center UC Davis on July 22 2010 The attendees are shown in Attachment 2 Summary of Discussion The discussion among
43. d system physics required required Models for black box models models are specific to a usually low computational requirements theoretical foundations widely documented system takes time to collect a training dataset requires a training set representing correct operation Table 4 Strengths and Weaknesses of FDD Models B 15 Refrigerant Quantitative Qualitative E ra gt Ki Timeseries Figure 1 FDD Approach Matrix Used to Classify FDD Approaches Each cell in this matrix will be studied in detail Most cells do not represent available tools but one can still envision what the characteristics of tools would be in these cells The sections that follow describe these hypothetical approaches in detail Where there are actual tools available they will be cited as examples and described By evaluating hypothetical approaches we can identify which would have the potential for appropriate tools and for Title 24 This will simplify the search for appropriate tools by allowing us to focus on a smaller set of tools 4 3 Description of Available and Developing Tools There are very few tools currently on the market A handful of other tools have been piloted but have not yet been introduced to the market as viable products and yet others are under development While we will discuss the approaches that can be taken to FDD in a generic hypothetical way it is useful to describe the to
44. e e Liquid temperature e kWh 2 Defined as the life cycle Net Present Value of the incremental benefit of the technology including the time dependent value of reduced energy consumption maintenance costs and quantifiable non energy benefits such as improved comfort divided by the sum of the incremental first costs and Net Present Value of all other incremental costs associated with the technology including maintenance costs and costs of remediation of faults B 7 Lovv Ambient temperature e Return air temperature Supply air temperature e Static Pressure e Suction pressure e Liquid pressure e Discharge Pressure e Power factor e Power quality Processing of diagnostic algorithms can take place in the onboard controller on an installed PC or remotely Even when a PC or remote computer is used there may still be a need for on site signal processing to reduce the data and pre process them In most cases these processing platforms do not contribute significantly to the cost For some methods however it will be significant High An approach that uses an EMS platform for processing Moderate An approach that that can be accomplished by an embedded controller Low An approach that can be accomplished only with use of an added PC or processor The defined scope for this program is remote diagnostics so all approaches considered here will require remote communications For remote diagnostics communications hardware and access
45. e or electrical and the type of model used for comparison vvith measurements first principles qualitative history VVe also identify the specific criteria that must be met to have a measure that is appropriate for inclusion in Title 24 These criteria include significant energy savings cost effectiveness prevalence of the fault being detected probability that the fault vvill be fixed reliability of detection deployability and other maintenance benefits From this foundation the team has developed draft standards for FDD in RTUS suitable for inclusion in California s Title 24 This draft is presented in the third section of the report C Draft FDD Standards for Rooftop Units An Industry Roundtable vvas held to get feedback on this draft and the results of this Roundtable are summarized here Keyvvords Rooftop unit air conditioning fault detection diagnostics Title 24 energy standards Please use the follovving citation for this report Heinemeier Kristin VVCEC Mark Cherniack NBI and Julien Bec UCD 2010 Fault Detection And Diagnostics Moving The Market And Informing Standards In California California Energy Commission CONTENTS A Interviews with Key Stakeholders B FDD Prioritization C Draft FDD Standard for Rooftop Units Public Interest Energy Research PIER Program PROJECT REPORT Fault Detection and Diagnostics Moving the Market and Informing Standards in California A Interviews with K
46. e of these units and their performance is suffering Most RTUs have some sort of fault that is increasing their energy use If these faults could be found and addressed then energy savings could be realized In Fault Detection And Diagnostics Moving The Market And Informing Standards In California Deliverables FDD Prioritization we identified nine different potential approaches depending on the type of data collected air side refrigerant side or electrical and the type of model used for comparison with measurements first principles qualitative history We also identified the specific criteria that must be met to have a measure that is appropriate for inclusion in Title 24 These criteria included significant energy savings cost effectiveness prevalence of the fault being detected probability that the fault will be fixed reliability of detection deployability and other maintenance benefits The faults that can be detected by various FDD tools include efficiency degradation low charge coil fouling filter dirty insufficient capacity excessive operating time incorrect control sequence lack of ventilation insufficient economizing unnecessary outdoor sir failed sensor control problems failed compressor stuck damper slipping belt leaking valves short cycling unit not operational The tools that are available are shown below Tool Name Status Data Model Developer FDSI Insight V 1 Available Refrigerant Quantitative Fiel
47. e usually already available The remaining hardware includes a transmitter which cost may vary according to the technology used Target price 100 5400 Example Available Tool Lo Cost SMDS The Low Cost Diagnostic Module proposed by Pacific Northwest National Laboratory is an example of a power monitoring tool that uses qualitative methods to diagnose system condition from power monitoring and simple airside measurements 5 Evaluation of Approaches The sections that follow describe how well the different FDD approaches meet the criteria just discussed and Table 8 below provides a summary of this assessment Recall that we can only provide a qualitative assessment of most of the criteria so we use a scale of High Moderate Low to describe whether each of the criteria are met fully partially or not at all Table 8 Summary of Evaluation of Different Fault Detection Approaches High O Moderate Low Sa So 5 2 5 T S a gt v 2 E 0 2 0 o E ot v lt Bee Bs s be lt 23 46 4 5 TE E 5 E E fo iz Savings e Cost Eff O Y O O Frequency 69 Fixed O Reliable O O O O Deployable O O O O Maint O 5 1 Magnitude of Energy Savings All approaches can offer significant energy savin
48. egradation Efficiency Degradation Not Economizing VVhen it Damper Not Modulating Should Excess Outdoor Air Faults shall be reported to a fault management application accessible by day to day operating or service personnel or annunciated locally on zone thermostats A performance indicator shall be provided vvhich vvill allovv tracking of efficiency The FDD System used shall be certified by the CEC and verified to be installed correctly Certification and verification procedures are TBD Attachment 2 Roundtable Participants Mike Brambley Pacific Northvvest National Lab Martha Brook California Energy Commission Cathy Chappell Heschong Mahone Group Mark Cherniack New Buildings Institute Bobby DiFulgentiz Lennox Industries Martyn Dodd EnergySoft Piotr Domanski National Institute of Standards and Technology John Douglas Lennox Industries Joseph Fleishman California Energy Commission Craig Fulgum Virtjoule Tom Garcia CalBO Sean Gouw Southern California Edison Dale Gustavson Better Buildings Institute Inc Kristin Heinemeier Western Cooling Efficiency Center Randall Higa Southern California Edison Sherry Hu Pacific Gas amp Electric Marshall Hunt Consultant John Kaufmann Pacific Northwest National Lab Golam Kibrya California Energy Commission John Kimmes Target David Kuo
49. emperature Based on that information the system can evaluate performance indices such as e Evaporation temperature Superheat e Condenser temp e Sub cooling e Evaporator Delta T e Condenser over ambient Example of detected faults e Faulty refrigerant charge Compressor valve leakage Liquid line restriction Condenser fouling e Inadequate airflow Hardware requirements amp Pricing elements e Temperature pressure sensors e data processing module in order to run the physical model e Communication module required for any FDD Example Available Tool Sentinel and Insight The Sentinel provided by Field Diagnostic Systems is an example of a refrigerant side tool that uses quantitative methods to diagnose system condition Due to the high cost of the Sentinel with its requirement for a number of sensors to be installed FDSI has developed the Insight analysis tool which makes use of onboard sensors and data acquisition systems to provide the same diagnostics Example Available Tool ClimaCheck Description This system developed in Sweden allows the continuous monitoring of refrigerant systems Available either in portable or fixed installation format the system is composed of the following elements A data acquisition system PA Pro e Power meter e Temperature sensors e Pressure sensors e Climacheck software e Optional LAN connection or GPRS modem The system is based on an internal thermo physical
50. er drier faulty transducer excessively cold return air or stuck open economizer when the ambient temperature is low T408 Dirty Filter Dirty Filter T414 Economizer Damper Actuator Out of Calibrate economizer E CAL If problem still exist then determine vvhat is Calibration limiting economizer rotation Economizer Damper Actuator Torque Above Load Limit Alert Actuator load too high Check damper load Economizer Damper Actuator Hunting Excessively Damper position changing too quickly Economizer Damper Stuck or lammed No economizer motion Check damper blades gears and actuator Economizer Damper Actuator Mechanical Failure Check actuator and replace if necessary Economizer Damper Actuator Direction Switch Wrong Actuator direction control switch CCW CW wrong Trane Intellipak Alarm Codes Alarm Blocked Air Return Energy Recovery Wheel Proof Failure Improper Airflow for Dehumid Low Pressure Control Open Ckt 1 Low Pressure Control Open Ckt 2 Low Refrigerant Charge Ckt 1 Low Refrigerant Charge Ckt 2 Cond Sump Min Level or Drain Failure CO2 Sensor Failure Min OA Flow SETPOINT Failure Min Position SETPOINT Failure OA Temp Sensor Failure Return Air Temp Sensor Failure Supply Air Temp Sensor Failure B 26 Trane Intellipak Alarm Codes Explained DIAGNOSTIC DISPLAYED REASON FOR DIAGNOSTI
51. ey Stakeholders A Report of the Western Cooling Efficiency Center WESTERN COOLING EFFICIENCY CENTER Prepared for California Energy Commission MARCH 2011 Prepared by Western Cooling Efficiency Center CEC 500 08 049 University of California Davis 1450 Drevv Ave Suite 100 Davis CA 95618 wcec ucdavis edu Introduction Objectives The objectives of the overall Fault Detection And Diagnostics Moving The Market And Informing Standards In California Deliverables project are To identify appropriate technology for identifying and diagnosing faults in commercial building Rooftop HVAC Units RTUs To develop a proposal for a standard that would require this Fault Detection and Diagnostics FDD for Title 24 To get industry feedback on this proposal To revise the proposal and submit it to the Title 24 process Approach Phase Phase 1 Obtain input from industry stakeholders on desired capabilities of FDD tools and service models for making best use of FDD tools Identify the faults that occur in RTUs and their impact and frequency to estimate the degree of savings made possible by FDD tools Prioritize those faults to determine which are most likely to be cost effectively diagnosed and addressed Identify diagnostic approaches that can be used to detect these faults Define a set of criteria for the attributes an approach must have to be likely to be successful in the market place or to be successfull
52. field studies all located in California and the North West This gives a very relevant set of data for our purpose These data are shown in Table 1 B 3 Table 1 Frequency of some Common Faults in Rooftop Air Conditioners Covvan 2004 Fault Fault Level Frequency Found Economizer Malfunction Various faults 6496 Refrigerant Undercharge Overcharge 5596 46 Inadequate Airflow lt 300 cfm ton 42 Failed Sensors Various faults 20 The information regarding the energy impact of each fault comes from Breuker and Braun 1998 Common Faults and their Impact for Rooftop Air Conditioners In this study different common faults were artificially introduced in an RTU and the impact on energy efficiency and COP was evaluated Table 2 shows the results that Breuker and Braun found Table 2 Energy Impacts of some Common Faults in Rooftop Air Conditioners Breuker and Braun 1998 Change in Change in Fault Fault Level capacity COP Compressor Valve Leakage 35 An 21 3 23 8 Condenser Coil Fouling 35 area blocked 21 3 23 8 Inadequate Airflow 36 A airflow 19 4 17 4 Liquid Line Restriction 20 AP 17 2 8 7 Refrigerant Undercharge Overcharge 14 undercharge 8 4 6 Economizer Malfunction Up to 40 Thermostat Errors Up to 40 Failed Sensor Up to 40 2 1 Mechanical Faults 2 1 1 Compressor valve leakage When slugs of liquid refrigerant enter the compressor the compressor valves can be damaged If the sealing becomes less
53. gs The Airside Quantitative approach provides somewhat lesser savings because it has a limited number of situations it can diagnose Most tools can identify refrigerant charge and airflow issues which may have the potential for energy savings Several methods can identify economizer problems and only one Hybrid Air Power Timeseries Based specializes in issues such as nighttime operation Depending on the cost it is likely that any of these methods would provide moderate or high energy savings and would be suitable for inclusion in Title 24 The energy savings may not be dependable enough to drive marketability however 5 2 Cost effectiveness The Power Qualitative and Hybrid Air Power Timeseries Based models would seem to have the most potential to be cost effective Since they require a very minimal set of data the cost of acquiring data would be quite small If they require significant hardware for processing however the costs will rise B 34 Most of the rest of the approaches vvould provide moderate level of cost effectiveness They vvould have to be analyzed carefully to determine vvhether or not they vvere cost effective The refrigerant based models because of their requirement for a large number of sensors can be expected to be less cost effective If these data can be obtained inexpensively for example by being mass produced and installed in the factory or by making use of existing sensors reliably as is done by FDSI s Insig
54. he benefit VV Long payback lack of public awareness Have to do the check hard to convince The customer is still comfortable so it s a hard sell e Hovv important is metering o o Car analogy mpg information very educative Need a monetary conversion process EER to Hard to do with energy efficiency A 5 VVhy vvould a customer replace a unit o Old and broken dovvn for example compressor out Age and cost of repair taken into account Under 5005 policy if it costs less than 500 repair it In a house 12 years and above compelling to replace it Another significant cause could be a hard to fix coil failure Window of opportunity with fluid change gt if some equipment is not on sale anymore need to change the whole unit More significant with split system because each half is sold independently If one needs to be replaced and only the new refrigerant type is available then the other needs replacement too many units in a complex for example Can be planned during the year Dismounted equipment used as spare for the rest of the units On RTU parts available for a long time Fluid available Concern for change if many RTUs at the same location and some are dead All may be replaced to avoid a mix of refrigerants Part of a building renovation Reroof ownership change tenant change Financial transaction already taking place Part of the negotiation to extend or renew the lease Depreci
55. he requested power profile of the unit is compared with target values stored into the controller for the different phases of the system Required information Current and voltage sensors at the RTU feed In case of three phase system two of each is necessary Example of detected faults Electrical electromechanical RTU faults Fault Loss of phase Locked rotor Slow starting motor Unbalanced voltage Short cycling Motor disconnect failure to start Incorrect control sequence Contactor improper contact closure Fan rotor faults that result in imbalance Non electrical faults Fault Refrigerant leakage undercharge or overcharge Loss of volumetric efficiency leaky valves seals Fouled condenser coil Dirty supply air filter Liquid ingestion 1 COP degradation 2 Diagnostic Method Current and Voltage Start transient Start transient Voltage Event sequence Event sequence Event sequence Phase current interruption transient Amplitude spectrum in steady operation Diagnostic Method Change of mean Start transient Change of mean Change of mean Anomalous transient 1 requires refrigerant flow or compressor map and head pressure 2 requires air side flow and RH Hardware requirements Pricing elements e single phase diagnostic can be implemented in a low cost 150 MHz PC on a chip with a 2 channel A D converter with an incremental RTU manufacturing cost as low as 200 e Communication module required f
56. he value of some of the maintenance benefits are B 11 High Allovving contractors to send the right technicians vvith the right tools By diagnosing problems remotely the contractor can plan for a service call more accurately This saves time and travel costs by avoiding second trips Increasing uptime for customers many of whom would suffer great financial losses if the RTU were to go down A service that can avoid system failures would provide a great benefit to a Customer and a great differentiator for a Contractor who could charge a premium rate for this advanced service e Ensuring that maintenance or service is done correctly to avoid callbacks Unbillable callbacks are a real drain on contractors bottom line An example of this is systems that can ensure that charge has been adjusted correctly Moderate e Reducing the frequency of required maintenance by annunciating when maintenance is required While this would seem to be a benefit to service contractors those interviewed for this study indicated that they would not be likely to rely on this and would provide service calls on the usual schedule They might reduce unscheduled maintenance however e Allowing maintenance to be provided remotely By assessing the condition of a system remotely a service technician might be able to provide service to the customer over the phone or even to make changes in controls remotely This reduces unscheduled maintenance calls
57. heir energy use If these faults could be found and addressed then a significant energy savings could be realized This report summarizes the activities that fulfill Objectives 1 5 as described in Introduction to Project 5 3 It documents the results of a thorough literature search and a set of interviews that shed light on the faults that occur in RTUs and on the potential approaches to detecting these faults It starts by describing the faults that affect RTUs in terms of their frequency and their potential for energy savings A set of criteria are then proposed according to which the approaches are then evaluated It then describes and classifies various approaches to detecting faults providing examples of existing tools where available Finally a set of prioritized approaches is identified which will guide the later development of a set of minimum attributes of a tool that could be adopted into code 2 Faults that Affect RTUs In this section the different faults affecting RTUs will be described For these faults data regarding the prevalence and energy efficiency impact will be presented The information regarding the prevalence of faults on this paragraph is quoted from Cowan 2004 Review of Recent Commercial Roof Top Unit Field Studies in the Pacific Northwest and California This study gathers data from previously completed research projects Slightly over 500 units in 181 locations have been inspected during these different
58. ht and Sensus MI then these methods may have some promise 5 3 Frequency of Fault All of the approaches address faults that have a significant frequency of occurrence As described earlier most existing RTUs can be expected to have one of the problems that each of these approaches can detect 5 4 Probability that Fault will Get Fixed Because most of the methods can detect issues such as incorrect refrigerant charge and the need for filter replacement it is likely that the faults will be remedied once detected Those faults are not necessarily the most important and energy intensive however Economizer malfunctions and certain airflow restrictions which are a big source of energy waste do not cause comfort problems and are more costly to address 5 5 Reliability Robustness Because refrigerant data can be difficult to obtain accurately the Refrigerant Cycle approaches have only moderate reliability While Air Side data can also be difficult to obtain accurately Qualitative and Timeseries Based approaches which do not require the same degree of accuracy as Quantitative approaches might be considered somewhat more robust Power based approaches have somewhat less difficulty in installing and are considered reliable 5 6 Ease of Deployment The deployment path for timeseries based methods is more limited than for the other methods Because they require a lengthy period of correct operation for training and cannot be used ina Co
59. ince it only has one sensor and it can start immediately in detecting faults without the requirement for a training dataset Such a tool would provide a great deal of information to building operators and service contractors NILM developed by MIT is an example of this approach Hybrid Airside Power Timeseries Based Another high scoring approach is this one which uses a dataset consisting of air side and power data to identify a limited set of faults The faults include overall efficiency degradation capacity degradation operation during unoccupied times and excessive cycling While it cannot identify some of the faults that are identified by most of the other approaches addressing this set of faults can be expected to save quite a bit of energy The simplicity of this approach is appealing and it should be quite reliable Because it cannot be used in a commissioning like service it is considered less deployable and because the faults it detects are silent faults that do not affect comfort or equipment condition it may not be considered highly marketable The Low Cost SMDS proposed by PNNL is an example of this approach Power Timeseries Based This approach requires looking at power at relatively high sampling frequency in order to detect faults that are expressed during transients in addition to those found by looking at mean values t requires only one datapoint but high frequencies require a different processing platform increa
60. increase the probability that faults will be fixed Tools that diagnose problems that are likely to be fixed will be more likely to result in savings so they are more likely to be successful in being implemented in code High e Refrigerant leakage undercharge overcharge e Condenser coil fouling e Inadequate airflow e Liquid line restriction e Other controls problems e Locked rotor Moderate e Liquid slugging e Compressor valve leakage e Economizer misapplied e Economizer malfunction e Loss of phase e Unbalanced voltage e Flooded start e Short cycling B 9 3 5 Reliability Robustness Reliability and robustness in a FDD tool refers to a number of factors that influence whether a tool can be expected to vvork vvell over time Some of these factors are How difficult is it to install the tool Installation errors such as placing sensors in the wrong location or mounting them incorrectly can render a FDD tool useless How prone is the tool to misinstallation Airside sensors can be difficult to install correctly and care must be taken to mount them in the correct location so that they accurately reflect the temperature of an entire air stream Refrigerant sensors are somevvhat less difficult but they are also prone to mounting errors as when temperature sensors are not mounted in direct enough contact vvith the lines being measured e Does the tool require maintenance Some sensors such as RH sensors and Pressure gauge
61. lems Failed Compressor Stuck Damper Slipping Belt Leaking Valves Unit Not Operational Basic FDD x O Extended FDD Suction Pressure Discharge Pressure Suction Temperature Discharge Temperature Liquid Temperature OAT RAT MAT SAT Condenser leaving air temp OARH RARH MARH SARH Damper Position Operating Mode Compressor Status Fan Status Vibration Fan Compressor or RTU kWh Power factor Real reactive power Power Quality Compressor Voltage Compressor Amperage RTU Voltage RTU Amperage Table 6 Faults Detected by Available Tools 3 8 S a 2s B 2 Ti 28 2 o S ER 3 S a 8 Eg ce mn o o o o o o o o o o o o o o x o o o x o x x o o o o x o o o o o o o x o o o o o o o o o o o o o x o o x o o x o o o o o o x o o x o x x o o x o x o o o o o o O O o o o o o o o o o x o o o o o o o o o x o x o o o o o o Table 7 Required Data for Available Tools ME gt Y z S So 2 28 65 Z 2 2 5 Se Za 5 o 3 o EE ME 2 2 on O nm o ce 2 O O x x O x o o x x o O O o o O o o x x x O x O x x x x x x O O x
62. llation is customer signs a service contract Best arguments to get people enroll into a maintenance contract routine vs crisis Most agree but still need to be economically persuasive we re taking care of it for you more of a collaborative work ask you while planning yearly budgets etc e What are the most common problems o Belt problems bearings contactors capacitors fan motor compressor missing filters filter change lovv air freezups heat exchanger failure expansion valve takes out compressor refrigerant leaks operating pressures Overcharging is a training issue Depends on the quality of maintenance It s possible to put sensors on everything but would have more issues with the sensors than vvith the parts in the first place e Special notes on economizers o It is the first thing you disable Complexity vs benefits it costs 500 1000 to fix it If disabled everybody gets vvhat he vvants Possible because of cheap energy economizer working is quite complex Setting it correctly is a challenge not noticed by the customer Noticed during maintenance or customer complaining about low cooling Service technician fresh air over dramatized gt need for less fresh air ie less bills most of the time Customers often prefer to have it disabled than fixing it if the customer knevv the cost of disabling the economizer it could be possible to make a value proposition Right now they don t see t
63. lling to pay to have this tool available What types of customers do you believe would be most interested in this kind of service D process server room financial component is important Would you be interested in remote monitoring of RTUs dependent on customer size or would you like for all customers What performance monitoring FDD information do you think are most important in terms of optimizing performance efficiency D Service assistant type information performance efficiency Supply and return temps amps on components to confirm what s running Go back and look at trends What information would you like to see on an FDD PM GUl dashboard D Daily updates although there s a nuisance factor Safevvay for example gets Benchmark variance data monthly on different variables Need to knovv hovv to use Carrier alarms R Traiing manufacturer school if it s vvorthvvhile then teach what you learned If you could design an RTU performance monitoring system with remote access what would be included as features Do you have any customers who you think we should interview Kevin Napper Safeway A 19 Public Interest Energy Research PIER Program PROJECT REPORT Fault Detection and Diagnostics Moving the Market and Informing Standards in California B FDD Prioritization A Report of the Western Cooling Efficiency Center WESTERN COOLING EFFICIENCY CENTER Prepared for California Energy Commission MARCH 20
64. m the air side methods This approach is very close to the air side timeseries based model on both required information and the way this information is treated The main difference is that the state of the fans and compressor is obtained from the power measured at the RTU feed Required information Power meter e OAT TAT MAT temperatures Example of detected faults e Efficiency degradation by increases in the total power use given the outdoor air temperature e Degradation in capacity from Increase in on time per cycle for each specific outdoor air temperature Continuous operation without cycling at a lower outdoor air temperature or lower outdoor air enthalpy than previously observed e Operation during unoccupied times or incorrect schedule specification via power level indicating supply fan condenser fan and compressor are operating during times when the building or specific building zones are not occupied Excessive cycling indicated by compressor power cycling at a frequency higher than acceptable e Unit not operational zero power during conditions e g time of week and outdoor air temperature when the unit has historically operated Hardware requirements amp Pricing elements Power meter the small number of electric components in a typical RTU 2 to 4 makes the identification of the running components straightforward obviating the need to measure the reactive power The requested temperature sensors ar
65. make no warrant express or implied and assume no legal liability for the information in this report nor does any party represent that the uses of this information will not infringe upon privately owned rights This report has not been approved or disapproved by the California Energy Commission nor has the California Energy Commission passed upon the accuracy or adequacy of ACKNOVVLEDGEMENTS The support of the California Energy Commission Public Interest Energy Research program staff as vvell as the staff of New Buildings Institute is gratefully acknowledged In addition Matthew Tyler of PECI has provided a great deal of insight and help to this work PREFACE The California Energy Commission Public Interest Energy Research PIER Program supports public interest energy research and development that will help improve the quality of life in California by bringing environmentally safe affordable and reliable energy services and products to the marketplace The PIER Program conducts public interest research development and demonstration RD amp D projects to benefit California The PIER Program strives to conduct the most promising public interest energy research by partnering with RD amp D entities including individuals businesses utilities and public or private research institutions PIER funding efforts are focused on the following RD amp D program areas e Buildings End Use Energy Efficiency e Energy Innovations Small Grants
66. mmissioning type service they can only be deployed in an ongoing performance monitoring or maintenance service 5 7 Other Maintenance Benefits From a Title 24 perspective maintenance benefits are an added bonus All of the tools that identify regular maintenance issues such as charge and fouling will provide benefits to the team that maintains the building The hybrid method that identifies issues such as nighttime operation would provide less of a maintenance benefit 6 Conclusion By providing a comprehensive classification for available and hypothetical FDD tools we have been able to identify which approaches seem the most promising which available tools might be suitable for marketization or for inclusion in standards 6 1 Highest Priority Approaches Through this process vve have identified five approaches that appear to be most likely to be appropriate Povver Qualitative One of the high scoring approaches is the Povver Qualitative approach The povver profile of the unit is compared with target values stored into the controller for the different phases of the system This is able to detect a large range of refrigerant and airside problems in addition to electrical and controls faults This approach has many benefits It can be quite cost effective since it requires only one sensor and can detect a large number of faults that create energy waste in a large fraction of buildings It would be difficult to install incorrectly s
67. n intrusive because the hardware can be connected on the electric power supply to the house as opposed to using end use meters to every appliances in the house The only required sensors are current and voltage sensors on the system In looking at instantaneous real and reactive power and knowing that each load has its own resistance inductance and capacitance characteristics it becomes possible to single out each individual load In the case of packaged air conditioning units the loads are limited to the compressors and fans t is therefore easy to detect which component is running If we combine this information to one or two air temperature measurements it becomes possible to get an insight into the working conditions of the system and changes of efficiency or major faults occurrences can be detected Combining information on the power uses and outdoor air and return air temperatures changes in energy efficiency and occurrence of major faults can be detected B 32 4 4 10 Hybrid Povver Airside Timeseries Based Model Approach This approach focuses on reducing the cost of the device and therefore focuses on faults that can be diagnosed by looking at event sequences In this case the system can be very simple since there is no need to measure the reactive power and the sampling frequency can be also reduced target a few samples minute This approach is called hybrid because it also requires some temperature information fro
68. n status e Cooling heating mode operation e Outdoor air damper position Example of detected faults e Economizer errors or thermostat set point errors can be detected by comparing the mode of operation cooling heating fan on off damper position for similar OAT RAT MAT and schedule e Capacity degradation can be seen from a variation in compressor cycling for similar OAT RAT temperatures e Fouled filter or other airway obstructions can be detected by an increased mixed air temperature for a given mode of operation and return outdoor air temperatures Hardware requirements amp Pricing elements e Factory sensors information gt no additional sensor required e Data logger Simple signal processing hardware for process timeseries based approach e Communication module required for any FDD Example Available Tools Sensus MI Description This product is entirely software based and leverages on the information already available through the BAS system Cost Very low due to low capital and installation costs ROl lt 1 year 4 4 7 Povver quantitative model approach This approach necessitates running a complete model taking into account the fluid dynamics mechanic and electromechanical aspects of the system in order to calculate the instantaneous povver and required povver requested by the unit The level of complexity involved renders this approach impracticable B 29 4 4 8 Povver Qualitative Model Approach In this case t
69. nput has been obtained on the draft standard a formal Code Change Proposal will be developed This will be done in conjunction with Portland Energy Conservation Inc The role of the CEC team will be defining requirements and reviewing draft proposals This will be completed by April of 2011 or whenever the deadline is for submission of proposals Follow the CEC review process The interviews and analysis conducted by the CEC team are part of the due diligence that is required for any Code Change proposal This process also includes formal stakeholder workshops and an open comment period The CEC team will participate in these workshops and responding to questions as needed and appropriate This is expected to occur in April of 2011 C 6 Attachment 1 Proposed FDD Standard for RTUs 2011 Title 24 for Nonresidential Buildings requires installation factory or field verification and acceptance testing of a Fault Detection and Diagnostics FDD system for Packaged Direct Expansion Units as a prescriptive measure As with any prescriptive measure this measure can be traded off for another optional measure with equal savings Credit is given using the Alternative Calculation Method by degrading cooling efficiency by XX for non FDD systems and only 5 for FDD systems This measure supplements the compliance option on Form MECH 12A first included in 2008 FDD capabilities must be verified in the field by verifying that the FDD hardware is installed
70. nto a maintenance contract W routine vs crisis Most agree but still need to be economically persuasive we re taking care of it for you E more of a collaborative work ask you while planning yearly budgets etc Q advice for good deals E give the clients the information they need How long is a typical service call how many service calls can a technician make in a day What influences this W typically 5 6 calls a day Compressor replacement 4 hrs but simple maintenance and diagnostics 4 6 day Q productivity gains of integrated diagnostics VV better diagnostics on the system makes you faster and more confident What is the nature of RTU service marketing how is service marketed How is service delivered scheduled calls on demand only for some other How is service linked to product sales What tools do your techs use to diagnose faulty or less than optimal performance W Temperature probe gages amp meters So much variation out there that it is tough to have a standardized protocol tools Some art is involved What tools do the OEM s provide for diagnosing faulty operation W Some things provided on the controllers Mainly on the furnace side Not much on the cooling side Usually used by technicians if present but not very predominant right now What tools do the OEM s provide for performance monitoring Do your techs use these tools If not why not What related to your RTU service busines
71. ods available and contractor practices explain that this is a very common problem in the field In the documented sample 46 of units present a charge outside of a 5 acceptance range Energy savings impacts found by Breuker and Braun were up to about 5 for a 15 error in charge Other researchers looked at higher levels of charge error and found even greater energy impacts For example Robert Mowris recently tested a residential air conditioner and found upwards of 65 efficiency loss in a unit that was 40 undercharged Personal communication with Robert Mowris 2010 Regardless of the energy savings the greenhouse gas reduction potential of detecting refrigerant leakage is substantial 2 2 Controls Faults 2 2 1 Short cycling Nominal efficiency of air conditioning is only reached after a fevv minutes of operation due to thermal transients Since most systems operate with on off control at partial loads the unit will naturally cycle between On and Off phases If this cycling sequence is too short though the unit stops working before even reaching an acceptable efficiency with a great impact on energy consumption This problem can be due to a bad thermostat setting or oversizing 2 2 2 Economizer malfunction On average 64 of the units of the sample presented some kind of economizer failure or tuning problem Different faults can be observed e Broken frozen or missing drive system components Outside
72. ols are cost effective Because they require a long period of proper operation to train the system their deployment model is less obvious Sensus Ml is an example of this approach B 36 While we have characterized and evaluated generic and hypothetical approaches to FDD it is certainly possible that individual approaches can overcome some of the hurdles faced by their category of approach For example if the cost and complexity of refrigerant side measurements could be reduced they could be very good candidates for code inclusion FDSI s Insight is an example of an approach that may have promise if it can use existing sensors 6 2 Next Steps The next step described in the next section of this report is to define a set of capability requirements that will provide the basis for the code language This requirements document will be vetted with industry stakeholders in a Roundtable meeting Coming out of this meeting will hopefully be a consensus as to the appropriate level of requirements for Title 24 We will then work to draft the proposed code language ACM descriptions and forms and take this proposal through the CEC rulemaking process If we do not find sufficient appropriate tools we will still develop a set of requirements and hold a stakeholders meeting Again we will drive towards a consensus on what should be in Title 24 someday even though there are not tools that can meet these requirements We will then put together a pl
73. ols that are commercially available available in pilot status only or in the pipeline Tables 5 through 7 describe the type the faults detected by each and the required input data respectively Information is shown for Basic diagnostics that implemented in the standard version of each tool and Extended diagnostics that implemented in potential extended versions Each of these tools is described more later Table 5 Available Tools Tool Name Data Model Status Developer FDSI Insight V 1 Refrigerant Quantitative Available Field Diagnostics Inc Sentinel Insight Refrigerant Quantitative Beta Field Diagnostics Inc ClimaCheck Refrigerant Quantitative Available ClimaCheck Inc SMDS Air Qualitative Pilot Pacific Northwest National Lab Sensus MI Air Qualitative Available University of Nebraska NILM Power Qualitative Pilot Massachusetts Institute of Technology Low Cost NILM Power Timeseries Pilot Massachusetts Institute of Technology Virtjoule Power Timeseries Developing Virtjoule Inc Low Cost SMDS Air Power Timeseries Developing Pacific Northwest National Lab B 16 o Basic FDD x Extended FDD Lovv Airflovv Low High Charge Sensor Malfunction Economizer not Functioning Compressor Short Cycling Excessive Operating Hours Performance Degradation Insufficient Capacity Incorrect Control Sequence Lack of Ventilation Unnecessary Outdoor Air Control Prob
74. on module required for any FDD The SMDS works by constantly collecting data from sensors installed on the equipment to measure its performance and detect and diagnose problems with its operation The unit then sends the results wirelessly directly from each packaged unit to a network operations center where the data are stored securely and information on the condition of each packaged unit is made available on the internet The SMDS can be installed on nevv or existing packaged air conditioners and heat pumps This approach gets rid of the physical model of the quantitative approach and relies on the variation of parameters and predefined acceptable ranges for each parameter to detect a faulty operation This vvould be much simpler but is only possible during steady state operation Cost Around 2000 unit 200 to 1000 for installation Other The diagnostics that are provided in some high end RTUs are an example of qualitative models that make use of air side data For example the Carrier controller can provide an alarm if a filter is clogged and if an economizer is malfunctioning based upon various sensor readings and a qualitative model of variables such as high static pressures and no damper movement mean The tables that follow provide the alarm codes that may be useful for energy management for four high end RTUs Lennox Alarm Codes Alarm Problem Action 6 S27 Dirty Filter Switch This indicates
75. or any FDD 4 4 9 Povver Timeseries Based Model Approach Source Armstrong et al 2004 2006 This approach is based on NILM Non Intrusive Load Monitoring which is the measurement of power and reactive povver used by the unit In this case many faults are diagnosed by looking at transients which implies a relatively high sampling frequency of around 120 Hz This method differs from the qualitative povver approach since the target values are not entered as system parameters but are obtained from the first phase of operation Required information Current and voltage at the RTU feed In case of three phase system two of each is necessary Example of detected faults Electrical electromechanical RTU faults Fault Loss of phase Locked rotor Slow starting motor Unbalanced voltage Short cycling Motor disconnect failure to start Incorrect control sequence Contactor improper contact closure Fan rotor faults that result inimbalance Non electrical faults Fault Refrigerant leakage undercharge or overcharge Loss of volumetric efficiency leaky valves seals Fouled condenser coil Dirty supply air filter Liquid ingestion 1 COP degradation 2 Diagnostic Method Current and Voltage Start transient Start transient Voltage Event sequence Event sequence Event sequence Phase current interruption transient Amplitude spectrum in steady operation Diagnostic Method Change of mean Start transient Change of mean
76. perating on defaults Alarm 1D indicates the Outside Humidity Sensor is out of Range Alarm 1E indicates the Return Air Humidity Sensor is out of Range Alarm 1F indicates the IAQ Sensor is out of Range Alarm 25 indicates the unit is locked out due either 1 high duct static pressure or 2 a faulty duct static pressure sensor vvith an output that is too high Alarm 26 indicates the control has detected a Supply Air Temperature fault for Cooling Alarm 27 indicates the control has detected a Supply Air Temperature fault for Heating Alarm 28 indicates the control has detected a Minimum Economizer Position fault condition Alarm 29 indicates the control has detected a Space Temperature Alarm condition Alarm 2A indicates a fault vvith the Duct Static Pressure sensor reading a lovv pressure vvhen there should be pressure present Alarm 2B indicates the Hot VVater Freeze Stat has opened indicating a fault that could cause the coil to freeze B 25 Carrier Alarm Codes Alarm Description Probable Cause T110 Circuit A Loss of Charge Low refrigerant or faulty suction pressure transducer T126 Circuit A High Refrigerant Pressure An overcharged system high outdoor ambient temperature coupled with dirty outdoor coil plugged filter drier or a faulty high pressure switch T133 Circuit A Low Refrigerant Pressure Low refrigerant charge dirty filters evaporator fan turning backwards loose or broken fan belt plugged filt
77. puts if well formulated detailed models can also complex and computationally intensive significant development effort many required inputs some detailed models unlikely to be a solution of choice simplified models may be used Models simulate faulty operation not readily available for easier detection sensitivity to poor user allow the modeling of inputs transients well suited for data rich specific to a system or may prove a good environments process choice if quantitative simple to develop and difficult to find a complete approaches too apply applicable set of rules demanding transparent reasoning mostly for complex systems good for one time works under simplicity can be lot as new assessments uncertainties rules are introduced Qualitative ability to provide depends on the expertise of Models explanations for the developer diagnostics because relies on cause effect relationships some methods do not require special knowledge of the system and accurate data well suited if theoretical gray box models require a suitable where no models are good process knowledge and other method exists unavailable inaccurate statistic expertise lower cost than some well suited if training low performances outside of methods data is abundant cheap the training data range good for long term no previous large amount of data for all assessments Process 5 understanding of the expected operation modes Timeseries Base
78. r contractors are looking for what you missed D we try to do 2 1 2 calls day lt 10 tons R 5 calls if they re simple if its superheat pressure or economizer 3 4 per day What is the nature of RTU service marketing how is service marketed How is service delivered scheduled calls on demand only for some other How is service linked to product sales What tools do your techs use to diagnose faulty or less than optimal performance D worked with three tools in Air Care Plus They were hit and miss A lot of the reports didn t make sense lt works better with small equipment eg temperature probes Depends on ambient and return conditions A lot of cables and connections It s hard to be profitable Existing customers will ask why didn t you find it before It s good to use when doing compressor changeouts for refrigerant charge we bought 2 tools on our own What tools do the OEM s provide for diagnosing faulty operation The new Lennox RTU tells you when it went off on high head delta T on coil pressures of system What tools do the OEM s provide for performance monitoring 10 Do your techs use these tools If not why not 11 What related to your RTU service business keeps you up at night Hovv do you or vvould you use FDD 1 Are you or your organization currently using any kinds of tools like FDD D VVe use Service Assistant R vve use thermal imaging Can you imagine using FDD tools in your service
79. re used to predict the behavior of the system On the other end of the spectrum black box models rely only on data from the process itself and the resulting model may not have any physical significance Table 4 illustrates some of the pros and cons of the different methods Quantitative models Quantitative models are directly based on the physics of the process under consideration According to the degree of complexity they can either be steady state linear dynamic or non linear dynamic In this case for a given set of measured inputs temperature pressure it is possible to calculate the expected behavior and compare it to the measured performances of the system analytical redundancy The differences residuals can then be evaluated to detect any fault within the system Qualitative models Qualitative models can be based on a set of rules inferred from the physics or expert knowledge or qualitative physics instead of relying on quantitative mathematical relationships Since measurement techniques usually provide quantitative inputs temperature some preprocessing is often required to convert this information into qualitative inputs One of the methods that can be used is for example fuzzy logic and other kinds of classifiers In the case of a rule based modeling technique a set of if then else rules are derived from a priori knowledge In expert systems the rules are derived from insight knowledge or guidance of people wi
80. s keeps you up at night VV Nothing specific to RTUs More general business management trained workers customer interaction E noise transfer can be an issue but outside of this no problem VV automated report generation every x month sent to a specific person E time it with energy bills so that customers are more aware VV on big businesses numbers add up and it starts to be worth investing And image issues A 14 Hovv do you or vvould you use FDD 1 Are you or your organization currently using any kinds of tools like FDD No 2 Can you imagine using FDD tools in your service business How What s the business model How vvould you change your service offering VV could be a benefit E depends on hovv it is packaged Do not overload clients technicians vvith data Do not see any reduction of field trips need to factor in false alarm W am avoiding more calls or creating more calls How many false positives Each layer of complexity adds possible faults gt the increased maintenance may counterbalance all the energy benefits On a service contract false alarms is a cost for the contractor 3 Would you be more interested in hard faults degradation faults or fault prediction 4 Would you be more interested in monitoring fault detection or diagnostics 5 What value would it bring to you reduce the number of trips required reduce the length of a service call E No real reduction of trips expected
81. s out of range Check Sensor resistance should be between 830 ohms 200 F and 345 7Kohms 40 F If so check field unit wiring between Sensor and ECEM OA Temperature Sensor The unit is reading a signal Unit functions that are PAR An automatic reset Failure that is out of range for the disabled include occurs after the OA Temp Problem The Outside Air Outside Air Temperature a Low Ambient Compressor input returns to its allowable Temperature sensor input is sensor input on the RTM Lockout range In order to prevent out of range Temp lt 55 F or Temp 209 b The Outside Air Damper rapid cycling of the Diagnostic Check Sensor resistance F drivestominimumposition there is a 10 second delay should be between 830 ohms c On VAV units with SA Temp before the automatic reset 200 F and 345 7Kohms 40 Reset type selected as OA F If so check field unit Temp Reset the Reset type wiring between Sensor and reverts to NONE for the RTM duration of the failure B 27 4 4 6 Air Timeseries Based Model Approach n this case there is no need of a physical model and no maps thresholds or set of rules have to be defined the system logs the system state for different input conditions and defines patterns If at some point during the system life the behavior of the system does not match these patterns the system is considered faulty Required information e Outdoor air temp e Return air temp e Mixed air temp e Fa
82. s require periodic calibration Any tool that requires such maintenance will be less robust than tools with maintenance free sensors A tool is only useful to a customer or technician if its rate of false alarms is minimized as well as the rate of missed diagnoses False alarms are a serious problem in FDD tools and anything that generates unreliable alarms and causes alarm overload will not be used for long Similarly if a tool cannot be counted on to detect when a system is failing it cannot be relied upon to provide remote diagnostics This delicate balance must be found for every tool and every type of fault detected High No maintenance is required no installation required or false positives and negatives are both minimized Moderate Minimal maintenance required installation is easy to perform moderate level of false positives and negatives Low Significant maintenance required installation prone to errors high false positives or negatives 3 6 Ease of Deployment FDD tools can be deployed in many different ways Three deployment models are described Performance monitoring Ongoing optimization of the use of the unit for a given environment can be referred to as Performance Monitoring Many FDD tools can be deployed in this way Commissioning Commissioning refers to a process of evaluating if a newly installed or existing unit is performing as expected It is a one time intervention where a technician is on site so
83. servicing contractor What is the threshold definition for detecting faults that drive the degradation factor How low is too low Further prioritize the faults in severity and frequency There are other performance issues that are not on the list in the Strawman It s a complicated picture with the manufacturers entrepreneurs customers and utilities all having separate driving interests Engage the utilities Emerging Technology programs for proof of product and then potentially tap into utility incentive programs and then into Title 24 FDD is an enabling technology While we can t make anyone use the information that is or will be available from this type of monitoring the fact that the information exists drives the potential for market understanding and enables action Some factor must be developed to account for this indirect benefit Sensor accuracy persistence is unquestionably an issue and is being addressed by ASHRAE We need to get up to speed on this activity Pick the top four faults in the Strawman along with Performance Degradation indicator and that s enough Take this discussion national to bring in additional manufacturers of FDD products Need the FDD business case and volume business to catch industry attention RTUs can be split into those with electro mechanical controls or with microprocessor controls This dictates what is possible with sensing and communications Transmit the performance data alarms FDD off the
84. sing the cost It should be able to detect a long list of faults and save quite a bit of energy Because of its simplicity it will be quite reliable and it will provide a great deal of maintenance information to service providers The fact that it cannot be used in a commissioning like service makes it somevvhat less deployable The Low Cost NILM developed at MIT and the Power Analyzer being developed by Virtjoule are examples of this approach Airside Qualitative The variation in air side parameters is compared with predefined acceptable ranges to detect faulty operation This can detect important and common issues such as economizer faults refrigerant charge errors and coil fouling It is reliable due to the fact that the data inputs have less stringent accuracy requirements The only downside to this method is its cost effectiveness Care will have to be taken to ensure that tools in this category are cost effective The System Monitoring and Diagnostic System developed by PNNL is an example of this approach Airside Timeseries Based This system logs the system state for different input conditions and defines routine patterns When a fault occurs such as economizer errors charge errors or coil fouling it can be detected as a deviation from the pattern These models have the benefit of simplicity and ability to detect important and frequent faults They too have only moderate cost effectiveness and care will have to be taken to ensure that to
85. ss than 100 Can t reduce service calls to 2 year because of filters and belts e Would customers pay a 5 10 premium O Can see a premium of 10 for a very good field diagnostic device If they have a critical environment Need a more sophisticated customer What type of customers would be most interested o Maintenance contract customers have a long term vievv and do analysis on a vvhole building level Higher end vvorkers like doctors lavvyers vvould be most interested in this kind of service High end restaurants Retails vvith food safety issues process server room financial component is important A 7 Attachment 1 FDD Interview Guide What is your service business like 1 What constitutes RTU service in the field 2 How is your service business structured 3 What problems are the most common or the most important why causes of service calls 4 Have you ever told an owner that their RTU was on its last legs What symptom were you responding to What did the customer say What data did you need to make this call and convince the owner 5 How long is a typical service call how many service calls can a technician make in a day What influences this 6 What is the nature of RTU service marketing how is service marketed How is service delivered scheduled calls on demand only for some other How is service linked to product sales 7 What tools do your techs use to diagnose faulty or less
86. t before It s good to use when doing compressor changeouts for refrigerant charge vve bought 2 tools on our ovvn VVhat are the benefits to the customer A 6 o o Better interaction vvith the contractor Shovvs professionalism long term performance monitoring may help to prepare the client for heavier preemptive interventions e What would you like in a tool o an automated report generation every so often month sent to a specific person time it with energy bills so that customers are more aware Do not overload clients technicians vvith data An indication about performance another one about energy consumption Different information levels idiot light alarm if efficiency goes under a certain threshold efficiency for each unit Tool needs to be portable log all activity Daily updates although there s a nuisance factor Safevvay for example gets Benchmark variance data monthly on different variables Will these tools reduce the number of service calls ooo o Do not see any reduction of field trips need to factor in false alarm am l avoiding more calls or creating more calls Hovv many false positives Each layer of complexity adds possible faults gt the increased maintenance may counterbalance all the energy benefits On a service contract false alarms as a cost for the contractor The system must actually bring what it promises A good solution should cost le
87. th expertise in the field Another method is to derive the rules from a first principle approach implemented in a tree structure within the software Data gathered during the system operation allows navigating the decision tree and reaching a conclusion about the unit s behavior Models derived from process timeseries This approach is entirely data driven a known set of input and output data is fed to the system which then tries to find a mathematical relation between the two This is called parameter extraction Two different approaches coexist if these parameters have no physical meaning the approach is qualified as black box If the choice of parameters is based on a first principle approach and their value is obtained through linear regression from the set of data we speak about a grey box approach instead Various methods exist for both approaches The main difference between the two lies in their abilities to make predictions outside of the training data range 4 2 Definition of FDD Approaches VVe defined in the first part tvvo vvays of categorizing FDD systems by the type of data required and by the type of model used VVhen combining this information vve can define an FDD Approach Matrix where in which almost any possible FDD approach can be located in Figure 1 B 14 Type of Model Strengths VVeaknesses Suitability for FDD Quantitative based on sound physics most accurate estimators of out
88. ult priorities prevalence e Take the RTU FDD discussion national including the manufacturers and ASHRAE 90 1 In summary participants collectively took a useful step on the path toward increasing maintaining and controlling the energy efficiency of RTUs through FDD methods as a component of enhanced performance monitoring and performance measurement C 4 Key Takeaways summarized as offered at the end of the meeting a As an alternative first step explore an approach to first capture the data perhaps in the T stat before analysis methods are standardized How long to store data Who collects analyzes How does it become actionable information Move toward a 2011 T24 Prescriptive Measure submission it s possible but challenging t was noted by the manufacturers that the hardware is the easy part and by the FDD tool developers that we can t manufacture this in a big way tomorrow Partnering may be required to move this technology into the market quickly Initiate an FDD Challenge similar to the idea of the Western Cooling Challenge to pull the products into the market more quickly and at the same time work to condition the market for the products This might happen in cooperation with the Retailer Energy Alliance that US DOE supports Assess both a performance degradation factor and a performance index as a potential overall performance monitoring fault condition indicator to the RTU owner manager
89. umented 4 4 2 Refrigerant Qualitative Model Approach This approach is very similar to the previously described refrigerant quantitative approach as in both cases a set of performance indices is calculated In this case though the value of these indices is compared to tabulated target values and the difference observed for each of these indices is used in a rule based approach in order to identify the fault Required information Suction pressure can be substituted e Liquid pressure with additional Discharge Pressure temperature sensors e Suction temperature Liquid temperature e Ambient temperature e Return air temperature e Supply air temperature Based on that information the system can evaluate performance indices such as e Evaporation temperature Superheat e Condenser temp e Sub cooling e Evaporator Delta T e Condenser over ambient Example of detected faults e Faulty refrigerant charge e Compressor valve leakage e Liquid line restriction e Condenser fouling e Inadequate airflow Hardware requirements amp Pricing elements e Temperature pressure sensors e data processing module e Communication module required for any FDD 4 4 3 Refrigerant Timeseries Approach n this case no target for the performance indices is defined The target values are obtained from the acquisition taking place during the first phase of operation Required information e Suction pressure can
90. y implemented in California s Title 24 energy code Evaluate the potential approaches according to these criteria and identify potentially successful approaches and describe the currently available tools that utilize these and the Energy and Demand Savings potentially Phase III Create a Minimum FDD Capability Requirements document that summarizes and describes the requirements for a tool that could be incorporated in Title 24 including a description of the required functionality and an outline of the acceptance tests that would be required to document installed functionality Hold an Industry Roundtable to obtain industry feedback on this draft Phase IV Develop and submit a code proposal and follow the CEC process to support the proposal A 3 n the second Phase of this project Fault Detection And Diagnostics Moving The Market And Informing Standards In California Deliverables B FDD Prioritization we will identify and prioritize the faults that can be detected by a set of currently or shortly available diagnostic tools and will evaluate the available tools One crucial part of this prioritization is collecting intelligence from key stakeholders In this report we describe the process of developing an interview guide and carrying out a small set of interviews We summarize the interviews that were held as well as provide the detailed responses to our list of questions In the third Phase we will develop

PIER/NBI Report on FDD - Western HVAC Performance Alliance

Contents

Download Pdf Manuals

Related Search

Related Contents