Research Projects Overview

Faculty members and students in the Building Systems Program, perform research in the broad area of energy management. Most efforts focus on energy use in buildings. The work includes applied research in the analysis of energy systems as well as practical implementation of energy-efficient technologies. The following information gives summaries of a wide variety of research projects over the past ten years.

Development and Evaluation of Integrated Mechanical Systems in Supermarkets
Client: Electric Power Research Institute
Point of contact: Mukesh Khattar (650) 855-2699
Amount of Grant: $97,500
Date: September 1997 - June 1998
Principal Investigator: Michael Brandemuehl
Abstract: Supermarkets represent unique applications of heating, ventilating, air conditioning, and refrigeration (HVAC&R) systems. They are characterized by year-round refrigeration requirements, high space heating requirements, low sensible cooling but high dehumidification needs, and high ventilation airflow rates. These features are also present in the new large facilities that combine in supermarkets and retail stores in a single large building (up to 200,000 ft2). Working with EPRI, equipment manufacturers, electric utilities, and Wal-Mart, a novel water-loop system for integrating the mechanical equipment systems in these stores has been develop, built, and is being tested at a store in Moore, OK. The objective of this project is to further evaluate these types of integrated system to address the unique loads of supermarket applications.

Techniques for Measuring and Controlling Minimum Outdoor Air Intake in VAV Systems (RP-980)
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $95,121
Date: September, 1997 - January, 1999
Principal Investigator: Moncef Krarti, Michael Brandemuehl
Abstract: The goal of the project is to perform theoretical and experimental analyses to compare various techniques for measuring and controlling minimum outdoor air intake rates in variable-air-volume (VAV) systems. Based on the results of these analyses, a set of guidelines will be developed to assess the accuracy and the ease of implementation of each technique.
Project Results: Project is currenty underway.

Ventilation in Parking Enclosures
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $90,000
Date: September, 1996 - August, 1998
Principal Investigator: Moncef Krarti
Project Results: Project is currenty underway.

Cooler Floor Heat Gain in Refrigerated Structures (RP-953)
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $48,621
Date: April, 1997 - August, 1998
Principal Investigator: Moncef Krarti
Abstract: The project has two main objectives. The first goal is to develop and validate detailed models for ground-coupled heat transfer calculations suitable for refrigerated structures. The results will be used to provide insights on the reltive importance of various design parameters on cooler floor heat gains. The second goal is to develop simplified yet flexible design methods to calculate heat transfer between ground and refrigerated structures.
Project Results: Project is currenty underway.

Whole Building Diagnostician
Client: Pacific Northwest National Laboratory
Point of contact: Michael Brambley (509) 375-6875
Amount of Grant: $130,000
Date: January, 1996 - January, 1998
Principal Investigator: Jan F. Kreider
Abstract
Project Results:

Model Based Optimization System
Client: Pacific Gas & Electric Company
Point of contact: Steve Blanc (510) 866-5570
Amount of Grant: $156,000
Date: July 1996 - July 1997
Principal Investigator: Activity 1 - Jan F. Kreider, Michael Brandemuehl
Activity 2 - Moncef Krarti
Abstract: The objective of the project was to develop the design for (1) the software that will perform optimization of building energy systems, and for (2) software and hardware tools that are necessary to configure the optimization software for a specific building. The software is as adaptable and modular as possible and will builds on existing calculation abilities. The project involved considerable analysis of factors that affect the design requirements of building control systems, including the utility pricing structure, the presence of thermal storage systems (TES), and the ability to forecast loads, weather, and real-time utility prices.
Project Results:
Software specification delivered to client
Developed guidelines for forecasting requirements for TES system control. Results indicate that , except for systems with very large storage relative to chiller size, forecasts of 12-16 hours are sufficient and that there is little value in forecasting demand later in the billing period.
Uncertainty analysis for errors in modeling and forecasting was developed, providing practical guidelines for control strategies with real-world uncertainties.
Three graduate students supported.
Three technical papers in preparation.

Implementation of Optimal On-Line Supervisory Control of Cooling Plants Without Storage
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $138,000
Date: September 1994 - August 1997
Principal Investigator: Michael J. Brandemuehl, Peter Curtiss
Abstract: The overall objective of the project was to evaluate and further develop, as necessary, the on-line optimal supervisory control methods developed under previous ASHRAE research projects through an implementation in the US WEST Advanced Technology Center in Boulder, CO. JCEM implemented on-line monitoring and optimal control of system setpoints to minimize a selected energy cost function while satisfying system constraints. On-line parameter estimation techniques for learning the plant characteristics and for selecting optimal control setpoints were evaluated using measured plant performance. The complete on-line optimal control methodology was field tested to demonstrate robust operation in adapting and producing optimal system operation. The seasonal operating costs of the test building were compared for optimal and conventional control strategies through simulation.
Project Results:
25-35% savings from "recommissioning" building control system
Only about 7% savings from optimal control
Developed and implemented on-online optimization system using component-based models developed from measured data.
Developed guidelines for optimal setpoints using off-line optimization
One PhD student supported
Two technical papers

Methodology Development to Measure In-Chiller, Fan, and Pump Performance
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $91,000
Date: April 1994 - January 1996
Principal Investigator: Michael J. Brandemuehl, Moncef Krarti
David Claridge, Jeff Haberl, Texas A&M University
Abstract: The overall objective of the project was to develop methods for in-situ testing of chillers, fans, and pumps. The methods were designed to estimate the energy cost impact of improved energy efficiency of electric-powered HVAC equipment. Specifically, the project identified a set of short-term testing methods to characterize the actual (in-situ) performance and energy use efficiency of chillers, fans, and pumps, and developed techniques to extend the short-term results to estimate long-term performance. The methods were validated with field testing.
Project Results:
Demonstrated that long-term energy use and peak demand can be estimated from short-term data, on the order of 1-3 days.
Developed rigorous method for uncertaintly analysis, allowing uncertainty calculations to guide data acquisition needs.
Results are being incorporated into ASHRAE GPC 14.
One graduate student supported.
Three technical papers.

Optimal Control of Ice Storage Systems
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $60,500
Date: April 1994 - June 1995
Principal Investigator: Moncef Krarti, Michael Brandemuehl
Abstract: The objective of this research project was to develop and evaluate control strategies that minimize operating costs and energy charges in commercial buildings with ice storage. The objective was be achieved through computer simulation of a variety of representative commercial building energy systems, including such factors as ice storage system type, chiller type, building load profile, climate, and electric rate structure. Optimization methods and strategies were based on a combination of techniques, including direct search algorithms, dynamic programming and mixed integer programming.
Project Results:

Development, Analysis and Monitoring of Improvements in Dehumidification for Commercial Buildings
Client: Electric Power Research Institute
Point of contact: Mukesh Khattar (650) 855-2699
Amount of Grant: $393,239
Date: July 1991 - December 1997
Principal Investigator: Michael J. Brandemuehl
Abstract: The objectives of the project are to analyze alternative HVAC and dehumidification systems and operating strategies for dehumidification in commercial buildings. The project has been focused on the unique load requirements of supermarkets and retail stores, especially in the warm and humid climate of the southeastern U.S. The evaluation includes analysis of field data obtained from a retail store in Mississippi, and supermarkets in Florida, Massachusetts, Illinois, and New Jersey. Alternative strategies include reduced airflow, air bypass, coil temperature control, dual-path systems (direct conditioning of ventilation air for moisture removal), heat pipe heat exchangers, and advanced desiccant systems. Computer simulation is also used to complement and extend the results of the field tests to alternative types of HVAC equipment and strategies for system operation and control. Models of alternative HVAC equipment have been developed, as needed, for both stand-alone performance analysis and integration with the general supermarket energy simulation. Simulation tools are used to evaluate alternative HVAC and dehumidification system performance under various operating conditions to develop optimal operation and control strategies.
Project Results:
Over four years of field data collection and analysis.
Cooling and dehumidification loads have been characterized for supermarkets, showing extremely low load sensible heat ratios. The presence of refrigerated cases dramatically reduces sensible cooling requirements without the corresponding reductions in latent loads.
Airflow control and heat pipe heat exchangers have been shown to be effective at reducing energy consumption and improving indoor conditions.
Using validated models of supermarkets and DX cooling systems, optimal design and operation strategies have been developed.

Small Scale, On-Line Diagnostics for an HVAC System
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $59,000
Date: April 1995 - July 1996
Principal Investigator: Jan F. Kreider, Peter S. Curtiss
Abstract: The objective of the research project was to demonstrate the feasibility of applying on-line diagnostics for HVAC systems. While previous work relied on expert systems for diagnostic inference, this work involves the use of neural networks and probabilistic belief networks to identify faults for diagnosis and decision evaluation. The generic approach is developed specifically for the operation of a fan-powered mixing box in a variable-air-volume distribution system. The results were demonstrated on a full-scale HVAC system in the JCEM Larson Building Systems Laboratory.
Project Results:

Developed and implemented real-time diagnosis system in actual HVAC system for three different fan-powered mixing boxes.
Demonstrated value of probabilistic belief networks for fault detection without the need to impose faults for training.
One PhD student supported
Several technical papers under development

A Prototype Expert System for Commercial Building HVAC Diagnostics
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $55,000
Date: April, 1994 - August, 1996
Principal Investigator: Jan F. Kreider
Abstract: A project to assist building operators to identify a class of HVAC problems is being prepared. It is restricted to small commercial building systems.

Time-Varying And Steady-State Heat Transfer From Slabs And Basements
Client: ASHRAE
Point of contact: William Seaton
Amount of Grant: $105,000
Date: 1992 - 1994
Principal Investigator: Moncef Krarti
Abstract:
1. A general solution for the steady-state heat conduction problem under a slab-on-grade floor with horizontal insulation is presented. The soil temperature field, the heat flux along the slab, and the total slab heat loss are obtained and analyzed using the Interzone Temperature Profile Estimation (ITPE) technique. The derived solution addresses all the common configurations for horizontal insulation of slab-on-grade floors. The effect of the outer inner edge insulation on heat flux variation along the slab floor surface and on total slab heat loss is discussed and analyzed. Finally, the influence of water table level on total slab heat loss is illustrated for various inner edge insulation configurations.
2. Steady-state temperature field distribution beneath vertically insulated slab is derived using the Interzone Temperature Profile Estimation (ITPE) technique. A water table is considered at a finite depth below the soil surface. The heat flux variation along the slab is discussed as well as the effect of vertical insulation length and value on the total heat losses from the slab-on-grade floor. It is shown that when the depth and/or the thermal resistance of the vertical insulation increases, slab heat loss decreases following the law of diminishing returns.
3. A general solution for the steady-periodic heat conduction problem under a slab-on-grade floor with horizontal insulation is presented. The soil temperature field, and the total slab heat loss are obtained and analyzed using the Interzone Temperature Profile Estimation (ITPE) technique. The derived solution addresses all the common configurations for horizontal insulation of slab-on-grade floors. A parametric analysis is conducted to determine the effect of thermal insulation U-value and length on the mean, amplitude, and phase lag of total slab heat loss.
4. The steady-periodic temperature field distribution beneath a vertically insulated slab is derived using the Interzone Temperature Profile Estimation (ITPE) technique. A water table is considered at a finite depth below the soil surface. The temperature variation within the ground is discussed as well as the effect of vertical insulation length and thermal resistance on the total heat losses from the slab-on-grade floor. It is shown that when the depth and/or the R-value of the vertical insulation increases, the annual mean and amplitude of total slab heat loss decreases while the phase lag generally increases following the law of diminishing returns.
5. The Interzone Temperature Profile Estimation (or ITPE) technique is applied to determine the steady--state temperature distribution within earth in contact with a partially insulated rectangular basement. A water table at constant temperature is assumed to exist at a given depth below the soil surface. The derived solution addresses all the common insulation configurations for basements and crawl spaces. The effects of insulation configurations, geometric dimensions and water table temperature on the heat flow mechanism near basements are discussed. The developed solution is compared with the Mitalas method to predict yearly average basement heat losses. The two methods are found to be in good agreement.
6. The Interzone The Interzone Temperature Profile Estimation (or ITPE) technique is applied to determine the steady--periodic temperature distribution within earth in contact with a partially insulated rectangular basement. A water table is assumed to exist at a given depth below the soil surface. The solution derived addresses all the common insulation configurations for basements and crawl spaces. The effect of basement depth and of partial insulation length and U-value on the mean, amplitude, and phase lag of total heat loss from basement floor and walls is investigated through parametric analysis.
Project Results:
Collection of test data
Creation of an analytical model for a wide variety of earth-to-building heat transfer
Creation and verification of time-varying model for soil temperature
Six technical papers
Two MS students supported

Neural Networks Applied to LoanSTAR Data
Client: Texas Energy Experiment Station
Point of contact: David Claridge
Amount of Grant: $10,000 + $15,000
Date: June - August, 1991; June, 1993 - May, 1994
Principal Investigator: Jan F. Kreider; Moncef Krarti
Abstract: Neural networks (NNs) were applied to the problem of predicting future energy usage given historical values. Electrical, heating and cooling energy needs on a Texas university building were predicted on an hourly basis to within a few percent accuracy. Several subsidiary studies were completed including evaluating the best NN architecture and learning parameters. The second study addressed the question of identifying system and load characteristics separately. In addition, predictions were sought farther into the future - for up to a year from the end of the training data set.
Project Results:
Five technical reports
Two conference presentations
One graduate student supported
Use of neural networks for the first time were demonstrated to very successful in predicting hourly building energy usage without immediately past consumption data being available

Colorado Renewable Data Network - WTHRNET
Client: Colorado Office of Energy Conservation
Point of contact: Robert Westby (303) 894-2144
Amount of Grant: $134,000
Date: January, 1987 - June, 1994
Principal Investigator: Jan F. Kreider
Abstract: Building designers, agribusinesses and energy planners require accurate weather data for a number of applications. The National Weather Service has an inadequate network of stations in Colorado with equipment requiring frequent calibration. The JCEM installed a new eight-station network during the summer of 1987 with sensors and local computers at each station designed for maximum reliability. The network has operated successfully for more than seven years with very high data integrity, greater than 96%. The data have been made available by paper or magnetic copy or by a bulletin board system. Users have included agribusinesses, telecommunications industries, utilities, federal agencies and private citizens.
Project Results:
Eight-station network installed state-wide
Development of new instrument for determination of both beam and diffuse radiation
Three technical papers
Six MS students supported
Monthly, two-page contribution to Colorado Climate publication
96% data integrity

Heat Loss And Moisture Condensation In Wall Corners
Client: Self funded
Point of contact: JCEM
Date: 1994
Principal Investigator: Moncef Krarti
Abstract: In this work a detailed, analytical, two-dimensional steady-state thermal analysis of insulated square corners is presented. The Interzone Temperature Profile Estimation (or ITPE) technique is employed, to obtain heat flow and temperature distributions for configurations of square structures with an insulation layer placed at the inside surface of the structure envelope. The problem of moisture condensation at the inner surface of the envelope is quantified by determining the extent of the envelope area that is susceptible to condensation. The results showed that even though corner heat loss is reduced with increasing insulation thermal resistance, the risk of moisture damage increases.
Project Results:
One technical paper

Design of a Soft Energy Community for Japan
Client: Japan Research Institute
Point of contact: Kenzo Tsutsumi
Amount of Grant: $45,000
Date: April, 1993 - December, 1993
Principal Investigator: Jan F. Kreider and Phillip Tabb
Abstract: Soft energy - renewable energy, energy conservation and clean energy systems - were applied to the design of a prototype soft energy community (SEC) in the Fukuyama City area near Hiroshima. Site energy including renewables, geothermal, waste energy and energy conservation were evaluated analytically. The available energy mix was matched to the needs of this community using the criteria of first and second law efficiency. A graphical and textual final report was prepared along with several renderings and energy flow diagrams
Project Results:
Final and interim reports on energy resources and end use matching
Several large color renderings of the community
Energy site plan

Commercial Building HVAC Controls
Client: Honeywell, Inc.
Point of contact: Gideon Shavit (708) 797-4184
Amount of Contract: $100,000
Date: September, 1992 - December, 1993
Principal Investigator: Michael Brandemuehl and Jan F. Kreider
Abstract: Commercial building control design is improved by laboratory testing proposed configurations in realistic and repeatable system conditions in a laboratory. The JCEM lab was used for validating a CAD system for control software and hardware design.
Project Results: Over forty separate applications (control system configurations) were tested and reported

Wind Energy Ice-Making
Client: National Renewable Energy Laboratory
Point of contact: Larry Flowers (303) 384-6910
Amount of Contract: $27,800
Date: September, 1993 - May, 1994
Principal Investigator: Michael Brandemuehl
Abstract: The objective of the project was to analyze ice-making applications for wind-hybrid power generation systems. The main applications are expected to be in remote communities of developing countries, where refrigeration is required for medical supplies and preservation of fish and produce. Direct-coupled systems offer the advantages of low cost, simplicity, and higher efficiency than systems with inverters. However, direct-coupling delivers power with a variable frequency, ranging from 30 - 100 hz, which could cause complications for ice-making equipment. The analysis included modeling and testing of two ice-makers that were directly coupled to variable-frequency wind turbine generators. Models were developed from measured data and used to evaluate potential applications.
Project Results:
Laboratory and field test results demonstrated proof-of-concept for direct-coupled wind and ice-making equipment.
Two technical papers.
One graduate student supported.

Residential Energy Conservation Studies
Client: Green Technologies, Inc.
Point of contact: John Hay (303) 581-9600
Amount of Contract: $25,000
Date: September, 1991 - December, 1993
Principal Investigator: Michael Brandemuehl
Abstract: A number of devices for reducing residential electrical demand were tested. They demonstrated good potential for demand reduction. This ongoing program is evaluating new devices continually.
Project Results:
Measured electrical demand reductions on home appliances
Measured electrical demand reductions on home electric motors

Control and Performance of Centralized Heating and Cooling Systems
Client: National Science Foundation (Award Number MSM-9021092)
Point of contact: Kenneth Chong
Amount of Grant: $50,000
Date: 1991
Principal Investigator: Michael J. Brandemuehl, Jan F. Kreider
Abstract: The project was a collaborative effort among the University of Wisconsin, University of Colorado, Johnson Controls, Inc., and the National Institute for Standards and Technology. (For administrative purposes, separate contracts were negotiated with the University of Wisconsin and the University of Colorado.) The focus of work at the University of Colorado was 1) to evaluate existing proposals for regression-based on-line optimal control of central plant heating cooling systems as proposed by Braun et al. (1989) as described above, and 2) to develop and evaluate neural network-based methods for the same task. The regression and neural network models were evaluated using measured data from the JCEM HVAC Systems and Controls Laboratory. A training set of tests, performed over a wide range of system operating conditions, was used to develop the models, and a second set of tests was performed to evaluate the ability of the models to predict performance at other operating conditions. Further testing was performed to evaluate complete HVAC system optimization using a neural network trained on system operating data for realistic diurnal and seasonal loads. Daily load profiles, representative of cooling loads in a two-zone building for the months of April through September, were imposed on the HVAC system and the measured data was used to train a system of neural networks.
Project Results:
Both the regression and network models fit well with the training data, giving RMS errors in power use of 2.5 kW and 2.7 kW, respectively, or approximately 4% of full power.
When the trained models were used to predict performance at other operating conditions, the RMS error for the regression model grew to 3.2 kW while that of the network model was reduced to 2.3 kW.
The neural network optimizer was shown to reduce HVAC system energy consumption by 15% over fixed-setpoint control.
Several technical papers have been presented through ASHRAE and ASME.

Capacity Credit of Grid-connected Photovoltaic Systems
Client: JCEM
Amount of Grant: Self funded, $15,000
Date: February, 1990 - August, 1992
Principal Investigator: Jan F. Kreider
Abstract: For decades solar systems have been considered to be fuel savers only in both thermal and electrical applications. For the first time, this study showed for a wide variety of climates that grid connected PV systems also have significant capacity credit. That is, they are able to replace conventional generation capacity for large utilities. This was demonstrated using the standard IEEE load duration curve in a number of distinct climates. Specifically, for grid penetrations up to 25% the capacity displacement has been shown to be of the same order as the average PV field output. The entire project was carried out analytically with results checked by hourly simulation. The analytical tools developed allow any utility to evaluate the capacity credit of PVs (and by extension, other renewables) quickly.
Project Results:
New, analytical method for evaluating the capacity credit of PVs
One graduate students supported.
Two technical papers, one technical report

Real Time Commercial Building Diagnostic Expert System
Client: Hewlett-Packard
Point of contact: William Reinert
Amount of Grant: $150,000
Date: June, 1989 - December, 1991
Principal Investigator: Jan F. Kreider
Abstract: A comprehensive data acquisition system provided hourly data used by an expert system to diagnose equipment operational problems in advance of serious equipment failure. In this computer manufacturing plant a project was completed to install a diagnostic expert system with several hundred rules for managing the HVAC systems. Predictions of expected building energy consumption (prepared using neural networks) are compared to hourly measurements. Discrepancies between the two can indicate potential problems. An incipient failure of a 700-ton centrifugal chiller was discovered and repairs were made before serious damage occurred, for example. Electrical and thermal knowledge bases and hourly predictors were constructed.
Project Results:
Four technical papers
Two conference presentations
Four graduate students supported
Use of neural networks for the first time were demonstrated to very successful in predicting hourly building energy usage

A Toolkit for Secondary HVAC Systems
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $65,000
Date: June, 1989 - December, 1990
Principal Investigator: Michael Brandemuehl
Abstract: One method for controlling electrical demand is to operate building heating and cooling systems off peak and to store the resulting heating and cooling effect for later use during peak periods. In this project we monitored eight large, chilled water, commercial building storage facilities in the US and Canada. The results showed that of the four common designs, one has a distinct advantage over all others. This approach is also the most economical from a life cycle cost viewpoint. Our successful completion of this project was a result of more than 15 years of experience in precise measurement of commercial building HVAC system performance in the field.
Project Results:
Documented algorithms (in Fortran) for all significant secondary systems used in the HVAC industry for buildings
One technical paper
Three graduate students supported

Primary and Secondary Terms Analysis and Renormalization
Client: Solar Energy Research Institute
Point of contact: Dr. Kris Subbarao (303) 231-1000
Amount of Grant: $15,000
Date: June, 1989 - December, 1989
Principal Investigator: Jan F. Kreider
Abstract: A key difficulty with the reliable prediction of energy consumption in solar buildings is resolving the discrepancy between the as built building and the as designed building. To ameliorate this problem the PSTAR approach developed at the Solar Energy Research Institute uses a short term test to establish key building characteristics. The present work involved developing a computer code to analyze PSTAR data and to demonstrate the methodology. Solar energy effects are one of the key areas of measurement and analysis in this work.
Project Results:
Software package completed
One MS student supported
One technical paper

Commercial Building HVAC and Solar Laboratory - "The Colorado HVAC Laboratory"
Client: Colorado Office of Energy Conservation
Point of contact: Robert Westby (303) 894-2144
Amount of Grant: $250,000 Industrial Matching: $250,000
Date: January, 1988 - December, 1989
Principal Investigator: Jan F. Kreider and Michael Brandemuehl
Abstract: The US heating, ventilating and air conditioning (HVAC) and solar industries are in need of a state of the art, full scale laboratory for the test of commercial building HVAC and solar systems and their controls. With support from industry, the Colorado Office of Energy Conservation and the University of Colorado, the Joint Center for Energy Management has constructed such a laboratory. Research in the laboratory has been focussed on systems research which could not be conducted anywhere prior to this time. The laboratory also provides an exceptional educational facility for students in the Building Systems Program at CU. Solar energy work includes solar collector testing, automated PV testing and insolation instrument development.
Project Results:
Completed $500,000 research laboratory in March, 1989
Four technical papers
Two PhD students supported
Ten MS students supported
Five undergraduate research opportunity appointments

Industrial cosponsorship program initiated Project Title: Energy Analysis and Diagnostic Program
Client: Colorado Office of Energy Conservation
Point of contact: Robert Westby (303) 894-2144
Amount of Grant: $414,000
Date: January, 1987 - June, 1990
Principal Investigator: Jan F. Kreider
Abstract: Energy must be used with maximum efficiency in commercial buildings and industry to maintain productivity. The JCEM has instituted a program for automating the audit process to include renewable energy as part of the audit report in both of these sectors. A number of expert systems have been written for selecting energy conservation opportunities, for diagnosing HVAC system problems and for decision analysis. A large knowledge base has been assembled for each of the expert systems. In addition to the expert systems work at CU, several dozen audits have been conducted by our partners at Colorado State University.
Project Results:
Multiple expert systems for energy analysis of buildings including renewables
Three technical papers
Seven MS students supported
$2 million in energy and demand savings in audited facilities
Solar energy modeling on new Denver International Airport terminal complex

Field Test of Chilled Water Storage
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grant: $85,000
Date: June, 1986 - December, 1988
Principal Investigator: Jan F. Kreider
Abstract: One method for controlling electrical demand is to operate building heating and cooling systems off peak and to store the resulting heating and cooling effect for later use during peak periods. In this project we monitored eight large, chilled water, commercial building storage facilities in the US and Canada. The results showed that of the four common designs, natural stratification has the distinct advantage over all others. This approach is also the most economical from a life cycle cost viewpoint. Our successful completion of this project was a result of more than 15 years of experience in precise measurement of commercial building HVAC system performance in the field.
Project Results:
Eight field sites tested for one week each
One technical paper
One graduate student supported
Identification for the first time of the most advantageous design

Improvements to the Modified Bin Method for Building Energy Analysis
Client: ASHRAE
Point of contact: William Seaton (404) 636-8400
Amount of Grants: $25,000, $23,000
Date: June, 1985 - December, 1989
Principal Investigator: Jan F. Kreider
Abstract: The bin method is a fast method for predicting energy consumption of solar and non-solar buildings. It offers a number of well-established advantages over the usual simulation approaches. The first phase of this project compared the modified bin approach of ASHRAE to hourly simulation results from the DOE 2.1C code. It was found that the method works well except for massive buildings or for buildings with exceptionally large glazing areas.
Solar energy effects on building load calculations were the topic of the second phase study. A number of new results enabled the bin method to be applied to solar and massive buildings with confidence. A major new finding is a new analytical relationship between meteorological data: binned dry bulb temperature and mean coincident insolation. This enables one to predict solar data on all orientations of vertical walls and for horizontal surfaces for many sites in the world where only ambient temperature records exist.
Project Results:
Bin method verified and validated for the first time
Four technical papers
Four MS students supported

Real-time Expert Systems for HVAC and Solar System Diagnostics
Client: Colorado Office of Energy Conservation
Point of contact: Robert Westby (303) 894-2144
Amount of Grant: $80,000
Date: February, 1988 - June, 1990
Principal Investigator: Jan F. Kreider
Abstract: Expert systems are computer programs which reason like human experts using the accumulated knowledge from a group of experts. For the past five years CU has been the leader in the use of expert systems software for HVAC and solar diagnostics. The present project resulted in the installation of a state of the art diagnostic expert system on a large manufacturing plant in Colorado. To accomplish this, interviews were held with experts, statistical predictors of energy consumption were developed and an expert system was written.
In addition, an expert system for diagnosing malfunctions in solar water heating systems has been written. Approximately 200 rules form the knowledge base. Graphics help and probabilistic reasoning trees enable the user to diagnose all known problems in standard solar water heating systems accurately. (Solar expert system written by John Klima.)
Project Results
First real time, diagnostic expert system installed on a building in the US
First interactive expert system for solar water heating system diagnostics
Three graduate students supported.
Two technical papers
New method for statistical analysis of building energy data

New Photovoltaic Based Device for Beam Insolation Measurement
Client: Colorado Office of Energy Conservation
Point of contact: Robert Westby (303) 894-2144
Amount of Grant: $10,000
Date: February, 1988 - June, 1990
Principal Investigator: Jan F. Kreider
Abstract: Beam insolation data are very sparsely collected in the US. However, these data are critical to the accurate prediction of PV system performance. However, present instruments for their measurement are costly both initially and over time owing to maintenance required for the tracking hardware. The JCEM has designed and field tested a stationary device capable of measuring all components of insolation without the need for tracking. The prototype tested showed that a very strong correlation existed between measurements of this device and beam insolation measured over the six-month test period by a traditional pyrheliometer.
Later work involved using neural networks to improve the accuracy of this device still further.
Project Results:
Development of a non-tracking device for insolation measurement.
Field test of the prototype over six months.
One graduate student supported.
Three technical papers
New method for beam radiation prediction

Stand Alone Photovoltaic System Design
Client: JCEM
Amount of Grant: Self funded, $38,000
Date: February, 1987 - June, 1992
Principal Investigator: Jan F. Kreider
Abstract: The design of stand alone PV systems has been done in the past using the laborious approach of making thousands of computer simulations to find the optimal PV/storage configuration to assure a required loss of energy probability. The design cost of such an approach became a significant fraction of the total system cost. The new approach uses applied stochastic process theory to predict the loss of load probability based on weather statistics which can be found readily from historical records. Systems can be designed preliminarily using this new approach without resorting to former inefficient simulation approaches. The next phase of the work to improve the methodology is now under way.
Project Results
New, analytical design method for stand alone PV systems
Two graduate students supported.
Two technical papers, one technical report

Test of a Matrix Solar Central Receiver
Client: Sandia National Laboratories
Point of contact:
Amount of Grant: $5,000
Date: February - May, 1987
Principal Investigator: Jan F. Kreider
Abstract: During the period of this contract a JCEM graduate student tested and analyzed a novel central receiver fabricated by Sandia Laboratories. The receiver was designed to be an air heater of high efficiency at high temperature. A simulation model was used as a design aid prior to the prototype fabrication.
Project Results
Fabrication of ceramic, matrix
Two technical papers
Two MS students supported