The work described in this report was done as part of a Cooperative Research and Development Agreement (CRADA) between the U.S. Department of Energy’s Pacific Northwest National Laboratory (PNNL) and KGS Building LLC (KGS). PNNL and KGS both believe that the widespread adoption of automated fault de4tection and diagnostic (AFDD) tools will result in significant reduction to energy and peak energy consumption. The report provides an introduction, and summary of the various tasks performed under the CRADA. The CRADA project had three major focus areas: 1) Technical Assistance for Whole Building Energy Diagnostician (WBE) Commercialization, 2) Market Transfer of the Outdoor Air/Economizer Diagnostician (OAE), and 3) Development and Deployment of Automated Diagnostics to Improve Large Commercial Building Operations.

This paper evaluates the cooling efficiency improvements that can be achieved by integrating radiant cooling, cool storage, and variable-speed compressor and transport motor controls. Performance estimates of a baseline system and seven useful combinations of these three efficient low-lift inspired cooling technologies are reported. The technology configurations are simulated in a prototypical office building with three levels of envelope and balance-of-plant performance: standard-, mid- and high-performance, and in five climates. The standard performance level corresponds to ANSI/ASHRAE/IESNA Standard 90.1-2004, Energy Standard for Buildings Except Low-Rise Residential Buildings (ASHRAE 2004a). From the savings estimates for an office building prototype in five representative climates, estimates of national energy saving technical potential are developed. Component and subsystem models used in the energy simulations are developed in a companion paper.

Component and subsystem models used to evaluate the performance of a low-lift cooling system are described. An air-cooled chiller, a hydronic radiant distribution system, variable-speed control, and peak-shifting controls are modeled. A variable-speed compressor that operates over 20:1 speed range and pressure ratios ranging from one to six is at the heart of the chiller. Condenser fan and chilled-water pump motors have independent speed controls. The load-side distribution is modeled from the refrigerant side of the evaporator to the conditioned zone as a single subsystem controlled by chilled-water flow rate for a specified instantaneous cooling load. Performance of the same chiller when operating with an all-air distribution system is also modeled. The compressor, condenser fan, and chilled-water pump motor speeds that achieve maximum coefficient of performance (COP) at a given condition are solved at each point on a grid of load and outdoor temperature. A variable-speed dehumidification subsystem is modeled and simulated as part of a dedicated outdoor air system to condition the ventilation air. A companion paper evaluates the annual cooling system energy use and potential energy savings to be gained by integrating radiant cooling, cool storage, and variable-speed compressor and transport motor controls.

This paper discusses setpoint-control strategies for thermostatically controlled appliances (TCAs) in a competitive electricity market, with the electric water heater load used as an example. By varying the TCA thermostat settings, the TCA power consumption can be shifted from the high-price period to the low-price period to reduce the peak-load and energy cost. Economic benefits and impacts on distribution feeder load shapes when applying different setpoint-control strategies are studied.

This paper is the second of a two part review of methods for automated fault detection and diagnostics (FDD) and prognostics whose intent is to increase awareness of the HVAC&R research and development community to the body of FDD and prognostics developments in other fields as well as advancements in the field of HVAC&R. The first part of the review focused on generic FDD and prognostics, provided a framework for categorizing methods, described them, and identified their primary strengths and weaknesses (Katipamula and Brambley 2005). In this paper we address research and applications specific to the fields of HVAC&R, provide a brief discussion on the current state of diagnostics in buildings, and discuss the future of automated diagnostics in buildings.

Lack of or improper commissioning, the inability of the building operators to grasp the complexity controls, and lack of proper maintenance lead to inefficient operations and reduced lifetimes of equipment. If regularly scheduled manual maintenance or re-commissioning practices are adopted, they can be expensive and time consuming. Automated proactive commissioning and diagnostic technologies applied to parts of the commissioning process address two of the main barriers to commissioning: cost and schedules. Automated proactive commissioning and diagnostic tools can reduce both the cost and time associated with commissioning, as well as enhance the persistence of commissioning fixes. In the long run, automation even offers the potential for automatically correcting problems by reconfiguring controls or changing control algorithms dynamically. This paper discusses procedures and processes that can be used to automate and continuously commission the economizer operation and outdoor-air ventilation systems of an air-handling unit.

An empirical or regression modeling approach is simple to develop and easy to use compared to detailed hourly simulations of energy use in commercial buildings. Therefore, regression models developed from measured energy data are becoming an increasingly popular method for determining retrofit savings or identifying operational and maintenance (O&M) problems. Because energy consumption in large commercial buildings is a complex function of climatic conditions, building characteristics, building usage, system characteristics and type of heating, ventilation, and air conditioning (HVAC) equipment used, a multiple linear regression (MLR) model provides better accuracy than a single-variable model for modeling energy consumption. Also, when hourly monitored data are available, an issue which arises is what time resolution to adopt for regression models to be most accurate. This paper addresses both these topics. This paper reviews the literature on MLR models of building energy use, describes the methodology to develop MLR models, and highlights the usefulness of MLR models as baseline models and in detecting deviations in energy consumption resulting from major operational changes. The paper first develops the functional basis of cooling energy use for two commonly used HVAC systems: dual-duct constant volume (DDCV) and dual-duct variable air volume (DDVAV). Using these functional forms, the cooling energy consumption in five large commercial buildings located in central Texas were modeled at monthly, daily, hourly, and hour-of-day (HOD) time scales. Compared to the single-variable model (two-parameter model with outdoor dry-bulb as the only variable), MLR models showed a decrease in coefficient of variation (CV) between 10 percent to 60 percent, with an average decrease of about 33 percent, thus clearly indicating the superiority of MLR models. Although the models at the monthly time scale had higher coefficient of determination (R2) and lower CV than daily, hourly, and HOD models, the daily and HOD models proved more accurate at predicting cooling energy use.

Refrigerant subcooling is a demonstrated and reliable way of increasing cooling capacity and system efficiency for conventional air-conditioning and refrigeration systems. There are several proven refrigerant subcooling devices available on the market; but this article will focus on a relatively new implementation which modifies a standard direct-expansion, vapor-compression refrigerant system through the addition of a liquid line heat exchanger downstream from the condenser and a mini-cooling tower to reject the heat from the heat exchanger. Because of additional first cost (heat exchanger and mini-cooling tower), it is economical only with certain applications.In general, the benefits of subcooling are higher in regions with high year-round temperatures (1,200 or more cooling degree-days to base 65 degreesF). Several of these new subcooling devices have been installed in federal facilities. Most installations are custom-designed to obtain optimum system performance, and data related to operation and maintenance are somewhat sparse. However, the technology is proving particularly applicable in direct-expansion, vapor-compression air-conditioning equipment, especially where old units are being replaced or where new construction/expansion or new installation is planned. It is not recommended as an add-on device.

Interest in combustion turbine inlet air cooling (CTAC) has increased during the last few years as electric utilities face increasing demand for peak power. Inlet air cooling increases the generating capacity and decreases the heat rate of a combustion turbine during hot weather when the demand for electricity is generally the greatest. Several CTAC systems have been installed, but the general applicability of the concept and the preference for specific concepts is still being debated. Concurrently, Rocky Research of Boulder City, Nevada has been funded by the U.S. Department of Energy to conduct research on complex compound (ammoniated salt) chiller systems for low-temperature refrigeration applications.

The performance was evaluated of a new US cooling technology that has been installed for the first time at a federal facility. The technology is a 15-ton natural gas-engine-driven rooftop air conditioning unit made by Thermo King. Two units were installed to serve the Navy Exchange at Willow Grove. The savings potential at Willow Grove is described and that in the federal sector estimated. Conditions for implementation are discussed. In summary, the new technology is generally cost-effective at sites where marginal electricity cost (per MBtu at the meter) is more than 4 times the marginal gas cost (per MBtu at the meter) and annual full-load-equivalent cooling hours exceed 2,000.

The 1994 Great Energy Predictor Shootout 11 (GEPS), sponsored by ASHRAE Technical Committees TC 4.7, Energy Calculations and TC 1.5, Computer Applications, involved modeling/predicting heating, cooling, and electric energy consumption in two large institutional buildings in central Texas (an engineering center and a business school building). This paper describes the methodology used by one of the winning GEPS entries.Energy consumption (E) in a large commercial building is a complex function ofclimntic conditions, building characteristics, building use, system characteristics, and type of heating, ventilating, and air-conditioning (HVAC) equipment used. Therefore, multiple linear regression (MLR) models and nonlinear modeling approaches, such as artificial neural network (ANN) models, tend to provide better modeling capabilities than simple linear regression modeling approaches (Katipamula et al. 1994; Kreider and Haberl 1994).The heating and cooling energy consumption in both buildings is modeled using the MLR approach. The electric energy consumption at both sites is weather-independent and is a function of only the building's operating schedule; there fore, it is modeled using a daytyping algorithm.At the engineering center the coefficient of variation (CV) for electric energy end-uses varied from 1% to 6%, while the CV for cooling energy consumption (E(c)) and heating energy consumption (E(h)) varied from 10% to 33%. At the business school building, the CV for electric energy end-use varied from 5% to 14%, while the CV for E(c) and E(h) consumption varied from 27% to 36%.

Regression models of measured energy use in commercial buildings are becoming an increasingly popular method of determining retrofit savings or identifying operational and maintenance (O and M) problems. When hourly monitored data are available, an issue that arises is what time resolution to adopt for regression models to be most accurate. This paper addresses this question by comparing monthly, daily, hourly, and individual hourly or hour-of-day (HOD) multiple linear regression (MLR) models when applied to measured cooling energy consumption ({dot E}{sub c}) in commercial buildings. {dot E}{sub c} consumption in five large commercial buildings in Texas (both under dual-duct constant-volume [DDCV] and dual-duct variable-volume [VAV] operation) is modeled in all four time scales using functional forms based on engineering principles. The relative advantages and disadvantages of all four types of models are discussed and compared. The outdoor dry-bulb and dew-point temperatures accounted for most of the variation (80% or more) in {dot E}{sub c}. Although the monthly models had higher model R{sup 2} than daily, hourly, and HOD models, the daily and HOD models proved more accurate at predicting {dot E}{sub c}. Also, the HOD models had higher model R{sup 2} and lower coefficients of variation (CV) than the hourlymore » models. The results of this study suggest that daily time scale models are most advantageous for retrofit savings determination, while HOD models are best for O and M purposes.« less

A proper understanding of the functional basis of energy use in residences has been crucial in spurring energy conservation in such building stock. In commercial buildings, however, such understanding is lacking, partly because there are numerous possible HVAC system types and control strategies, and partly because the detailed algorithms available to simulate energy use in various types of HVAC systems provide little insight into identifying ways of saving energy in existing buildings. The objective of this paper is to derive closed-form steady-state functional relations for air-side cooling and heating thermal energy use for four of the most widespread HVAC system types, namely terminal reheat and dual-duct, both under constant air volume and under variable air volume operation. Expressions are derived for hourly energy use as a function of climatic variables, building characteristics, and system parameters. The effects of economizer cycle and deck reset schedules are also treated. The expressions derived could be utilized to develop functionally accurate regression models of monitored energy use for retrofit savings determination and to ascertain whether the HVAC system is operating properly, as well as for parameter estimation from either short-term or long-term monitoring and for sensitivity analyses of how various physical and operating parameters affect energy use.

This report describes the results of a simulation of thermal energy storage (TES) integrated with a simple-cycle gas turbine cogeneration system. Integrating TES with cogeneration can serve the electrical and thermal loads independently while firing all fuel in the gas turbine. The detailed engineering and economic feasibility of diurnal TES systems integrated with cogeneration systems has been described in two previous PNL reports. The objective of this study was to lay the ground work for optimization of the TES system designs using a simulation tool called TRNSYS (TRaNsient SYstem Simulation). TRNSYS is a transient simulation program with a sequential-modular structure developed at the Solar Energy Laboratory, University of Wisconsin-Madison. The two TES systems selected for the base-case simulations were: (1) a one-tank storage model to represent the oil/rock TES system, and (2) a two-tank storage model to represent the molten nitrate salt TES system. Results of the study clearly indicate that an engineering optimization of the TES system using TRNSYS is possible. The one-tank stratified oil/rock storage model described here is a good starting point for parametric studies of a TES system. Further developments to the TRNSYS library of available models (economizer, evaporator, gas turbine, etc.) are recommended so that the phase-change processes is accurately treated.

HVAC systems of large commercial buildings consume energy in excess of the sum total of the building loads. This excess energy use is due to the fact that a single air-handler unit in a HVAC system, having to provide conditioned air at different supply temperatures to different zones in the building, can do so only by resorting to either (a) a certain amount of mixing of cold and hot air streams as in dual-duct systems or (b) to terminal reheating in single-duct systems. This mixing of cold and hot air streams or terminal reheating result in an energy penalty which can be minimized by say, converting a constant air volume (CAV) system to a variable air volume (VAV) system, but it cannot be entirely eliminated. This paper proposes an index, called the Energy Delivery Efficiency (EDE), which characterizes this penalty and rates the energy performance of HVAC systems on an absolute scale. We develop the mathematical basis of the EDE approach for both one-zone and two-zone buildings, that allows determining the variation of the ideal EDE with outdoor temperature for a specific building. Year-long measured whole-building cooling and heating energy use data from two retrofitted buildings are finally used to illustrate differences between actual EDE plots of CAY and VAV systems, how they compare with the ideal EDE of a two-zone building, and how the approach can provide diagnostic insights into HVAC system efficiency in specific buildings.

A nondimensional cycling model for estimating the seasonal performance of conventional air conditioners and heat pumps was developed. Starting with a simple time-constant model of the startup of an air conditioner, a nondimensional time variable was developed which captures three important influences of on-off cycling performance degradation: response of the system, fraction on-time of the system, and cycling rate. Experimental data from three air conditioners and one heat pump provided excellent agreement with the model. The model should be applicable to any cooling technology that has similar on-off characteristics to conventional air conditioners and heat pumps.

The retrofit of dual-duct constant volume systems (DDCV) with energy-efficient variable air volume systems (VAV) has become common in recent years. In general, the energy savings from such retrofits are estimated by developing a temperature-dependent regression model using whole building preretrofit energy consumption data. Model predictions are then compared with measured post retrofit consumption, to determine the savings. In cases where the preretrofit energy consumption is not available such a method cannot be implemented. This paper describes a method that can be used to calculate savings in such cases. The method is based on use of simplified calibrated system models. A VAV model was developed based on the ASHRAE TC 4.7 Simplified Energy Analysis Procedure (SEAP) (Knebel, 1983) and calibrated with the postretrofit energy consumption of a large engineering center in Central Texas. The loads from the calibrated VAV model were then used with the DDCV model to estimate the preretrofit energy use, also based on TC 4.7 SEAP, and apparent savings were determined as the difference between the DDCV predicted consumption and measured energy consumption for the postretrofit VAV system. The simulated hourly cooling energy consumption from the VAV model was within ±1GJ (±20 percent) of the measured consumption. The simulated daily consumption (the sum of 24 hours of consumption) compared better with the measured daily consumption (within ±7 percent). The apparent saving from the retrofit of the DDCV system with VAV was about 684 GJ in cooling energy and 324 GJ in heating energy for a three-week period June–July 1991.

An expression for the part load factor (PLF) of a heat pump operating in the cooling mode was developed based on a series of experimental tests conducted according to the standard heat pump test procedures. The tests included: cycling rates from 0.8 to 10 cycles per hour, fraction ON-times of 0.2 to 0.9, indoor dry-bulb temperatures between 22.2 °C and 26.7 °C (72–80 °F), indoor relative humidities between 20% and 67%, and outdoor temperatures between 27.8 °C and 37.8 °C (82–100 °F). The expression was based on the functional relationship of each independent variable (fraction ON-time, cycling rate, indoor dew-point and dry-bulb temperatures, and outdoor temperature) with the dependent variable (PLF). The PLF evaluated from the expression compared well with the PLFs from the experimental tests. Finally, the seasonal coefficient of performance (SCOP) estimated from evaluating the PLF from the developed expression was compared with SCOPa based on the ASHRAE/DOE test procedure. SCOPa was about 7% higher. SCOP for several climatic regions was computed and compared with the SCOPa. The difference between SCOP and SCOPa was between 6% and 20%.