Patent Grant
10732657
This invention relates in general to the field of energy management, and more particularly to a demand response prediction system that employs fine-grained energy consumption baseline data.
One problem with resources such as electricity, water, fossil fuels, and their derivatives (e.g., natural gas) is related to supply and demand. That is, production of a resource often is not in synchronization with demand for the resource. In addition, the delivery and transport infrastructure for these resources is limited in that it cannot instantaneously match production levels to provide for constantly fluctuating consumption levels. As anyone who has participated in a rolling blackout will concur, the times are more and more frequent when resource consumers are forced to face the realities of limited resource production.
Another problem with resources such as water and fossil fuels (which are ubiquitously employed to produce electricity) is their limited supply along with the detrimental impacts (e.g., carbon emissions) of their use. Public and political pressure for conservation of resources is prevalent, and the effects of this pressure is experienced across the spectrum of resource providers, resource producers and managers, and resource consumers.
It is no surprise, then, that the electrical power generation and distribution community has been taking proactive measures to protect limited instantaneous supplies of electrical power by 1) imposing demand charges on consumers in addition to their monthly usage charge and 2) providing incentives for conservation in the form of rebates and reduced charges. In prior years, consumers merely paid for the total amount of power that they consumed over a billing period. Today, most energy suppliers are not only charging customers for the total amount of electricity they have consumed over the billing period, but they are additionally charging for peak demand. Peak demand is the greatest amount of energy that a customer uses during a measured period of time, typically on the order of minutes. In addition, energy suppliers are providing rebate and incentive programs that reward consumers for so called energy efficiency upgrades (e.g., lighting and surrounding environment controlled by occupancy sensors, efficient cooling and refrigeration, etc.) in their facilities that result in reductions of both peak and overall demand. Similar programs are prevalent in the water production and consumption community as well.
Demand reduction and energy efficiency programs may be implemented and administered directly by energy providers (i.e., the utilities themselves) or they may be contracted out to third parties, so called energy services companies (ESCOs). ESCOs directly contract with energy consumers and also contract with the energy providers to, say, reduce the demand of a certain resource in a certain area by a specified percentage, where the reduction may be constrained to a certain period of time (i.e., via a demand response program) or the reduction may permanent (i.e., via an energy efficiency program).
The above examples are merely examples of the types of programs that are employed in the art to reduce consumption and foster conservation of limited resources. Regardless of the vehicle that is employed, what is important to both producers and consumers is that they be able to understand and appreciate the effects of demand reduction and efficiency actions that are performed, say, on individual buildings. How does a building manager know that the capital outlay made to replace 400 windows will result in savings that allow for return of capital within three years? How does an ESCO validate for a contracting regional transmission operator (e.g., Tennessee Valley Authority) that energy efficiency programs implemented on 1,000 consumers will result in a 15 percent reduction in baseline power consumption?
The answers to the above questions are not straightforward, primarily because, as one skilled in the art will appreciate, weather drives consumption. Weather is not the only driver in consumption, but it is significant. For instance, how can a building's energy consumption in January of one year be compared to its consumption in January of another year when average temperatures in the two months being compared differ by 25 degrees? Is the difference between the two month's power consumption due to weather, or implementation of an energy efficiency program, or a combination of both?
Fortunately, those in the art have developed complex, but widely accepted, normalization techniques that provide for weather normalization of energy use data so that consumption by a building in two different months can be compared without the confusion associated with how outside temperature affects energy use. These modeling techniques provide for normalization of energy use data for buildings and groups of buildings, and they are accurate for the above purposes when employed for energy use periods typically ranging from years down to days. That is, given sufficient historical energy use data (“baseline data”), a model can be developed using these normalization techniques that can be used to accurately estimate the energy consumption of the building as a function of outside temperature. These estimates are used to remove weather effects from an energy use profile and also to predict energy use as a function of temperature.
The present inventors have observed, however, that conventional normalization techniques, utterly fail to be accurate and useful when energy use data granularity is less than a 24-hour period. Normalization models that are derived from energy use data having granularities on the order of six hours, one hour, 15 minutes, etc., have been shown to be exceedingly deficient in accuracy and are thus unreliable.
Accordingly, what is needed is a technique that provides for accurately estimating energy use as a function of temperature, where the technique is derived from and is applicable to, energy consumption periods less than 24 hours.
What is also needed is an apparatus and method for employing fine-grained (i.e., less than 24 hours) energy use data to derive an accurate model for energy use based upon outside temperature.
What is additionally needed is a fine-grained baseline energy data weather normalization apparatus and method.
What is further needed is a system for characterizing a building's energy consumption as a function of temperature that is applicable at resolutions less than one day.
What is moreover needed are mechanisms that understand and employ the transient energy use responses of buildings for purposes of energy consumption predictions covering individual buildings, groups of buildings, and larger areas.
The present invention, among other applications, is directed to solving the above noted problems and addresses other problems, disadvantages, and limitations of the prior art by providing a superior technique for predicting a demand response event using an energy consumption baseline having a much finer granularity than that which has heretofore been provided. In one embodiment, a demand response dispatch prediction system is provided that includes a building lag optimizer, a dispatch prediction element, and a dispatch control element. The building lag optimizer is configured to receive identifiers for buildings participating in a demand response program, and is configured to generate energy use data sets for each of the buildings, each of the energy use data sets having energy consumption values along with corresponding time and outside temperature values, where the energy consumption values within the each of the energy use data sets are shifted by one of a plurality of lag values relative to the corresponding time and outside temperature values, and where each of the plurality of lag values is different from other ones of the plurality of lag values, and is configured to perform a non-linear parabolic analysis on the each of the energy use data sets to yield corresponding non-linear parabolic model parameters and a corresponding residual, and is configured to determine a least valued residual from all residuals yielded, the least valued residual indicating a corresponding energy lag for the each of the buildings, and non-linear parabolic model parameters that correspond to the least valued residual, and where the corresponding energy lag describes a transient energy consumption period preceding a change in outside temperature. The dispatch prediction element is coupled to the building lag optimizer and to weather stores, and is configured to receive, for each of the buildings, outside temperatures, the corresponding energy lag, and the corresponding non-linear model parameters, and is configured to estimate a cumulative energy consumption for the buildings and is configured to predict a dispatch order reception time for a demand response program event. The dispatch control element is coupled to the dispatch prediction element is configured to receive the dispatch order reception time and is configured to prepare actions required to control the each of the buildings to optimally shed energy specified in a corresponding dispatch order.
One aspect of the present invention contemplates a system for predicting a dispatch for buildings participating in a demand response program event. The system includes a building lag optimizer, a dispatch prediction element, and a dispatch control element. The building lag optimizer is configured to determine an energy lag for a building. The building lag optimizer has a thermal response processor and a non-linear parabolic analysis engine. The thermal response processor is configured to generate a plurality of energy use data sets for the building, each of the plurality of energy use data sets comprising energy consumption values along with corresponding time and outside temperature values, where the energy consumption values within the each of the plurality of energy use data sets are shifted by one of a plurality of lag values relative to the corresponding time and outside temperature values, and where each of the plurality of lag values is different from other ones of the plurality of lag values. The non-linear parabolic analysis engine is coupled to the thermal response processor and is configured to receive the plurality of energy use data sets and is configured to perform a non-linear parabolic analysis on the each of the plurality of energy use data sets to yield corresponding non-linear parabolic model parameters and a corresponding residual. The thermal response processor determines a least valued residual from all residuals yielded by the regression engine, the least valued residual indicating the energy lag for the building, where the energy lag describes a transient energy consumption period preceding a change in outside temperature. The dispatch prediction element is coupled to the building lag optimizer and to weather stores, and is configured to receive, for each of the buildings, outside temperatures, the corresponding energy lag, and the corresponding non-linear parabolic model parameters, and is configured to estimate a cumulative energy consumption for the buildings and is configured to predict a dispatch order reception time for the demand response program event. The dispatch control element is coupled to the dispatch prediction element and is configured to receive the dispatch order reception time and is configured to prepare actions required to control the each of the buildings to optimally shed energy specified in a corresponding dispatch order.
Another aspect of the present invention comprehends a method for dispatching buildings participating in a demand response program event, the method comprising: generating a plurality of energy use data sets for each of the buildings, each of the plurality of energy use data sets comprising energy consumption values along with corresponding time and outside temperature values, where the energy consumption values within the each of the plurality of energy use data sets are shifted by one of a plurality of lag values relative to the corresponding time and outside temperature values, and where each of the plurality of lag values is different from other ones of the plurality of lag values; performing a non-linear parabolic analysis on the each of the plurality of energy use data sets to yield corresponding non-linear parabolic model parameters and a corresponding residual; determining a least valued residual from all residuals yielded by the non-linear parabolic analysis, the least valued residual indicating a corresponding energy lag for the each of the buildings, where the corresponding energy lag describes a transient energy consumption period preceding a change in outside temperature; using outside temperatures, the non-linear parabolic model parameters, and energy lags for all of the buildings to estimate a cumulative energy consumption for the buildings, and to predict a dispatch order reception time for the demand response program event; and employing the dispatch order reception time to prepare actions required to control the each of the buildings to optimally shed energy specified in a corresponding dispatch order.
These and other objects, features, and advantages of the present invention will become better understood with regard to the following description, and accompanying drawings where:
Exemplary and illustrative embodiments of the invention are described below. In the interest of clarity, not all features of an actual implementation are described in this specification, for those skilled in the art will appreciate that in the development of any such actual embodiment, numerous implementation specific decisions are made to achieve specific goals, such as compliance with system related and business-related constraints, which vary from one implementation to another. Furthermore, it will be appreciated that such a development effort might be complex and time consuming but would nevertheless be a routine undertaking for those of ordinary skill in the art having the benefit of this disclosure. Various modifications to the preferred embodiment will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments. Therefore, the present invention is not intended to be limited to the particular embodiments shown and described herein but is to be accorded the widest scope consistent with the principles and novel features herein disclosed.
The present invention will now be described with reference to the attached figures. Various structures, systems, and devices are schematically depicted in the drawings for purposes of explanation only and so as to not obscure the present invention with details that are well known to those skilled in the art. Nevertheless, the attached drawings are included to describe and explain illustrative examples of the present invention. The words and phrases used herein should be understood and interpreted to have a meaning consistent with the understanding of those words and phrases by those skilled in the relevant art. No special definition of a term or phrase (i.e., a definition that is different from the ordinary and customary meaning as understood by those skilled in the art) is intended to be implied by consistent usage of the term or phrase herein. To the extent that a term or phrase is intended to have a special meaning (i.e., a meaning other than that understood by skilled artisans), such a special definition will be expressly set forth in the specification in a definitional manner that directly and unequivocally provides the special definition for the term or phrase.
In view of the above background discussion on building energy consumption and associated present day techniques employed to develop weather normalized energy consumption baselines that allow for analyses of building energy use, a discussion of the present-day techniques and their limitations and disadvantages will now be presented with reference to
Turning to
Going forward, energy consumption and associated discussions will employ terms corresponding to electrical energy usage (e.g., kilowatts, kilowatt hours) because electrical energy usage and derivation of electrical usage baselines are currently more prevalent in the art. However, the present inventors note that the principles and techniques disclosed herein according to the present invention are equally applicable to other forms of energy such as, but not limited to, water, natural gas, fossil fuels, and nuclear fuels.
Consider profile 101, where roughly 10 kilowatt hours (kWh) of electricity are consumed by the exemplary building during the weeks in January of the first year, decreasing down to a low of roughly 5 kWh during the weeks surrounding March of the first year, and increasing and peaking to slightly under 20 kWh during the summer months, decreasing to a low usage of roughly 5 kWh in the fall, and increasing up to roughly 10 kWh as it turns cold in the fall. Profile 101 is typical of many small- to medium-sized buildings (SMBs) in various locations in the world. These SMBs may comprise heating, ventilation, and air-conditioning (HVAC) systems to control climate within the SMBs at a comfort level supporting occupancy. Such is found in factories, schools, churches, airports, office buildings, etc. The HVAC systems may be very simple, and thermostat controlled, or they may be part of more complex building management systems (BMSs) that may include occupancy sensors, controlled lighting, and mechanisms to actively manage building energy use by varying activation schedules and/or duty cycles of equipment (e.g., compressors, evaporators, condensers, fans, lights, etc.).
Given that electrical energy is generally purchased from a utility provider (e.g. Tennessee Valley Authority), profile 101 implies that there are costs associated with heating the exemplary building that increase during colder weather and that decrease following the colder weather as seasonal temperatures increase. Profile 101 also indicates that there are costs associated with cooling the exemplary building that increase during warmer weather and that decrease following the warmer weather as seasonal temperatures decrease. But heating and cooling costs are not the only components of overall building energy consumption as is shown in profile 101. Other components may be due to energy use as a function of, but not limited to, occupancy, usage of large equipment, lighting, hours of operation, and equipment maintenance or malfunction issues. It is those other components of energy use that are of interest, in addition to energy use as a function of weather, to the present application. Yet, without installing costly and complex energy monitoring equipment within the exemplary facility, it is difficult at best to separate energy consumption as a function of the weather (i.e., “weather induced energy consumption”) from the other components of energy use.
To complicate matters, one skilled in the art will also appreciate that weather is not the same from year to year. Accordingly, profile 102 is presented as an example of energy consumption by the same exemplary building in a second year, where it is presumed that the configuration and use of the exemplary building may or may not be different in the first and second years. Yet, a building manager, utility grid manager, or energy service company analyst cannot discern the impact of other energy efficiency or demand reduction mechanisms on the building's energy use without first estimating the effects of weather on the building's energy use, subtracting those effects from the overall usage profiles 101-102, and normalizing those profiles 101-102 to a reference outside temperature (e.g., 65 degrees Fahrenheit (F)) so that the profiles 101-102 can be compared in a manner that will yield meaningful results. At this point, all that can be derived from the profiles 101-102 of
To address the uncertainties associated with determining how much energy an exemplary building should be consuming as a function of weather (namely, outside air temperature), those within the art have fielded standard techniques for estimating the effects of weather on a building's energy consumption, one of which is specified in Measurement of Energy and Demand Savings, ASHRAE Guideline 14-2002, published in 2002 by The American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. It is not the intent of the present application to provide an in-depth discussion of the different techniques for estimating weather induced energy consumption effects, for this will be evident to those of skill in the art. What is sufficient to note herein is the essence of these techniques and their limitations when employed to estimate weather induced energy consumption effects using fine-grained energy consumption data.
One skilled in the art will further appreciate that there are many applications for a baseline energy consumption model that is derived from coarse-grained or fine-grained energy use data, such coarse-grained data as is represented by the profiles 101-102 of
Now turning to
Accordingly, the profiles 101-102 of
Ê(i)=E(i)−M[T(i)]+M[TREF], where:
Ê(i) is an estimated weather normalized energy consumption for week i, E(i) is the actual energy consumption for week i, T(i) is the average weekly temperature for week i, M[T(i)] is the model estimate of energy consumption for the average weekly temperature T(i), TREF is a reference average weekly temperature, and M[TREF] is the model estimate of energy consumption for average weekly temperature TREF.
Thus, to normalize profiles 101-102 to 65 degrees, 65 is employed in the model 201 as TREF, yielding two energy use profiles that are normalized to 65 degrees, which can be compared or employed in other useful analyses. In other words, weather induced effects have been removed from the baseline energy use profiles 101-102 after normalization to 65 degrees.
The present inventors note, however, that the regression model 201 of
It is also noted that the granularity of baseline energy consumption data may be varied as well to develop a regression model. For example, rather than employing weekly energy consumption values to develop a baseline model for normalization purposes, monthly or daily values may be employed as well to provide insight into energy consumption of the exemplary building at a granularity that is required for a given analysis application.
The present inventors have observed, though, that the present-day modeling techniques discussed above with reference to
Referring to
Consequently, even though conventional weather normalization regression techniques have proved accurate and useful when employed to derive baseline models from coarse-grained energy use data (i.e., data with granularity equal to or greater than 24 hours), they utterly fail to yield model parameters that can be used to reliably and accurately estimate building energy consumption as a function of outside air temperature. This is a significant problem, for use of such a model, like model 301, to normalize fine-grained energy use data for purposes of comparison, estimation, or prediction, will result in gross error.
The present inventors have further observed that present day weather normalization techniques, such as those discussed above with reference to
The present invention overcomes the above noted limitations and disadvantages of the prior art, and others, by providing apparatus and methods for characterizing and creating accurate and reliable models of building energy consumption that are derived from fine-grained energy consumption data, namely, data obtained at intervals which are less than the energy lag of the building under consideration. For most SMBs, intervals on the order of one hour would otherwise result in the noisy distribution area 305 of
The present inventors have further observed that when energy consumption data is shifted in time relative to outside temperature data in a baseline at an amount approximately equal to a building's energy lag, and when regression analyses are performed on this shifted data, noisy distribution areas such as area 305 of
Referring now to
The baseline data stores 401 comprises fine-grained baseline energy consumption data corresponding to one or more buildings (or, “facilities”), where there is a sufficient amount of consumption data for each of the one or more buildings to enable an accurate energy consumption baseline regression model to be generated for each of the one or more buildings. In one embodiment, granularity of fine-grained baseline energy consumption data corresponding to some of the one or more buildings is one hour. In another embodiment, granularity of fine-grained baseline energy consumption data corresponding to some of the one or more buildings is 15 minutes. Other embodiments contemplate a combination of intervals that would be construed as “fine-grained” according to the present disclosure, such as 24-hour interval data for buildings having energy lags greater than 24 hours. Further embodiments comprehend fine-grained energy consumption data that differs in interval size from building to building within the stores 401. In one embodiment, the stores 401 may be collocated with the building lag optimizer 410 such as, but not limited to, within a network operations center (NOC) corresponding to an energy service company, an independent system operator (ISO), a regional transmission organization (RTO), a transmission system operator (TSO), or any of a number of other concerns that control and monitor operation of an electricity transmission grid. Other embodiments of the present invention contemplate deployment of the apparatus 400 within like-functioned facilities corresponding to control and monitoring of other energy sources as noted above. In all embodiments, the baseline energy consumption data for each of the one or more buildings comprises a time of day value or other type or value from which granularity of the baseline energy consumption data may be determined. Likewise, all embodiments comprise an outside temperature value corresponding to each data point of energy consumption for each of the one or more buildings.
In operation, the regression engine 412 functions to generate energy consumption baseline regression models characterized by model parameters for one or more fine-grained baseline energy consumption profiles, as are discussed above. Embodiments of the regression engine 412 comprehend a 5-parameter multivariable regression model that minimizes its residual term, or a 4-parameter multivariable regression model that minimizes its residual term, or other multivariable regression techniques that are known in the art for use in developing baseline energy consumption models. In one embodiment, the regression engine 412 may comprise a combination of the above noted regression models.
Baseline energy consumption data for a selected one of the one or more buildings may be downloaded to the thermal response processor 411 or the consumption data may be streamed over a network of interconnections known in the art. In addition, the thermal response processor 411 may be configured to accept downloaded or streamed data for a plurality of the one or more buildings simultaneously and may be employed to control the regression engine 412 for purposes of determining an optimal energy lag for one or each of the plurality of the one or more buildings whose baseline energy consumption data are obtained via the stores 401. For purposes of clarity, operation of the thermal response processor 411 will be discussed with reference to generation of an optimum energy lag and associated optimum regression model parameters corresponding to a single one of the one or more buildings.
Responsive to baseline energy consumption data that is received from the baseline data stores 401, the thermal response processor 411 provides the data to the regression engine 412 over LAGDATA along with a first value on THERMLAG that indicates an amount of time lag to shift energy consumption data relative to time stamp and outside temperature values in the baseline data. The first value on THERMLAG may be a time of day or may merely be an integer value indicating how many increments to shift energy consumption data so that it lags the time and temperature values in the baseline data by that number of increments.
Upon reception of the baseline data on LAGDATA and a first lag value on THERMLAG, the regression engine 412 performs a regression function as noted above to generate first model parameters and a first residual for a first corresponding model to be employed for analysis purposes. The first model parameters are output to bus OPTMPAR and the first residual is provided to the response processor 411 on bus RESIDUAL.
In a second iteration, the processor 411 generates a second value on THERMLAG that results in a corresponding shift in the energy consumption data relative to the time stamp and outside temperature values of the baseline data. The second value, in one embodiment, is an increment of the first value. Responsively, the regression engine 412 generates second model parameters and a second residual for a second corresponding model to be employed for analysis purposes. The second model parameters are output to bus OPTMPAR and the second residual is provided to the response processor 411 on bus RESIDUAL.
The iterations of energy lag continue, with generation of respective model parameters, lag values, and residuals until a lag threshold has been reached indicating that the baseline energy consumption data has been shifted a number of increments greater than an estimated energy lag of the building being modeled. In an embodiment that is using energy baseline data in 1-hour increments, 24 iterations may be performed, thus generating 24 thermal lag values and 24 sets of regression model parameters (e.g., parameters A-E in
Upon completion of the iterations, the thermal response processor 411 compares all of the residuals generated by each of the above iterations and determines which one of the residuals is less than all of the other residuals. The lag value that corresponds to the minimum value residual resulting from all of the above iterations corresponds to the energy lag of the building under consideration, and that lag value is output on bus OPTLAG along with model parameters that were generated using that lag value.
The present inventors note that other embodiments of the building lag optimizer 410 contemplate variations of the thermal response processor 411 that perform shifting of the baseline consumption data itself relative to time stamp and temperature data, and that receives model parameters from the regression engine 411, and which generates both values on OPTLAG and OPTMPAR. Via such embodiments a conventional regression engine 412 may be employed as opposed to one that performed the additional functions of time shifting the energy consumption data.
The building lag optimizer 410 according to the present invention is configured to perform the functions and operations as discussed above. The optimizer 410 may comprise logic, circuits, devices, or application programs (i.e., software) disposed within a non-transitory medium such as a hard disk or non-volatile memory, or a combination of logic, circuits, devices, or application programs, or equivalent elements that are employed to execute the functions and operations according to the present invention as noted. The elements employed to accomplish these operations and functions within the building lag optimizer 410 may be shared with other circuits, logic, etc., that are employed to perform other functions and/or operations commensurate with intended application.
Now turning to
According to the present invention, a lag LAG.0-LAG.N−1 equal to the time shift is recorded, and multiple regression model parameters MPAR.0-MPAR.N−1 and residuals RESID.0-RESID.N−1 are generated by the regression engine 412. The thermal response processor 411 then compares all N residuals and selects the one having the least value as the optimum residual. Accordingly, the lag value and model parameters corresponding to the optimum residual are designated as the optimum energy lag and optimum regression model parameters for the building under consideration. Henceforth, when analyses are performed for the building under consideration, the optimum regression model parameters and optimal energy lag are employed to perform weather normalization comparisons, efficiency analyses, consumption predictions, validations, etc.
The techniques discussed above with reference to
The present inventors also note that multiple variable regression analysis techniques are presented above with reference to the present invention in order to teach relevant aspects using prevalently known mechanisms in the art. However, it is noted that the present invention may also be embodied within configurations that utilize techniques other than multiple variable regression analysis in order to derive modeling parameters that accurately characterize a building's energy consumption and energy lag. Such techniques may include, but are not limited to, as one skilled in the art will appreciate, more than one statistical technique may be used to produce an approximation of a building's dependent energy relation with weather and lag. Any such technique is formalized as a function of independent parameters that describe baseline energy consumption data for the building relative to outside temperature and unknown mutually exclusive shifts in time of the baseline data's parameters. The difference between observed values of energy consumption and estimated values of energy consumption associated with such a function is a quantity that one skilled in the art seeks to minimize in order to improve model accuracy. Although the residuals are employed herein as a measure of goodness-of-fit, the present inventors note that other variables are contemplated by the present invention which include, but are not limited to, linear models of more or less parameters, non-linear models of a parabolic or higher polynomial order as well as machine learning modeling techniques (e.g., neural-networks, decision trees, etc.).
In addition to the embodiments discussed above with reference to
Referring now to
The facility stores 601 comprises identification data corresponding to one or more buildings to allow for characterization data to be associated therewith, and to allow for selection of corresponding baseline energy consumption data that is stored in the baseline data stores 604. In one embodiment, the baseline data stores 604 comprises fine-grained baseline energy consumption data corresponding to the one or more buildings as described earlier, and functions in substantially the same manner as the baseline data stores 401 described above with reference to
Operationally, the characterization system 600 is employed to determine characterizing features of the one or more buildings which include, but are not limited to, energy lag and optimal multiple regression baseline model parameters as are described above. The facility processor 602 retrieves facility data from the facility stores 601 for a selected building and provides this data to the lag optimizer 604 on FACDATA2. Responsively, the lag optimizer 603 retrieves one or more sets of energy consumption baseline data corresponding to the selected building from the baseline data stores 604 and performs the functions described above to generate an energy lag for the building along with optimal model parameters. The energy lag is provided to the facility processor 602 on bus OPTLAG and the optimal model parameters are provided on bus OPTMPAR. The facility processor 602 may subsequently select a second building from the one or more buildings and provide its data to the lag optimizer 603 for generation of a second energy lag and second optimal model parameters. The facility processor 602 may subsequently perform these functions for remaining buildings in the facility stores 601. In one embodiment, the facility processor 602 and building lag optimizer 603 may perform the disclosed functions serially for each of the one or more buildings. In another embodiment, the facility processor 602 and lag optimizer 603 may perform the disclosed functions for a plurality of the one or more buildings concurrently.
In one embodiment, the facility processor 602 may provide the energy lags and optimum model parameters for corresponding ones of the one or more buildings to the facility stores 601 via FACOPT, where the energy lags and optimum model parameters are stored and may be henceforth accessed for employment in other applications.
In another embodiment, in addition to providing the energy lags and optimum model parameters for corresponding ones of the one or more buildings to the facility stores 601 via FACOPT, the facility processor 602 may create categories of buildings that have been optimized according to the above. The categories may correspond to a common energy lag (e.g., 1-hour buildings, 2-hour buildings) or they may correspond to ranges of energy lag (e.g., fast buildings (0-4 hours), nominal buildings (5-8 hours), slow buildings (8-12 hours), etc.). The facility processor 602 may provide these categories as well over FACOPT so that buildings having similar energy lags may be identified and their parameters accessed for further application.
Advantageously, the system 600 according to
Turning now to
The dispatch processor 702 is coupled to a building lag optimizer 703 via a second facility data bus FACDATA2, an optimal lag bus OPTLAG, and an optimal model parameters bus OPTMPAR. The building lag optimizer 703 is coupled to a baseline data stores 704.
The participant stores 701 comprises identification data corresponding to one or more buildings that participate in the demand response program to allow for selection of corresponding baseline energy consumption data that is stored in the baseline data stores 704, to allow for energy lags to be associated therewith, and to further allow for employment of the energy lags in development of a schedule for dispatch control for each of the participants in the program event. In one embodiment, the baseline data stores 704 comprises fine-grained baseline energy consumption data corresponding to the one or more buildings as described earlier, and functions in substantially the same manner as the baseline data stores 401 described above with reference to
Operationally, the demand response dispatch system 700 is employed to determine energy lags of the one or more buildings that participate in the demand response program corresponding to the dispatch order. The system 700 also generates a dispatch schedule and performs dispatch for each of the one or more buildings to optimally shed the energy specified in the dispatch order in a timely manner. In one embodiment, the buildings with the highest values of energy lag are dispatched upon commencement of the program event because these buildings are presumed to exhibit a longer transient energy consumption response to changes in outside temperature, and thus they may exhibit a longer transient internal temperature response to abrupt changes in energy consumption, which may achieve demand response program event objectives while preserving comfort levels for internal occupants. Subsequent dispatches are performed in order of decreasing energy lags. One advantage of prioritizing dispatches in the noted order is that the effects of the load shedding for high energy lag buildings may not affect comfort of the occupants therein. The present inventors note, however, that if duration of the program event is longer than energy lags of some of the buildings participating in the program event, comfort levels may be affected.
To generate the above prioritized dispatch schedule, the dispatch processor 702 retrieves facility data from the participant stores 701 for a selected building and provides this data to the lag optimizer 703 on FACDATA2. Responsively, the lag optimizer 703 retrieves one or more sets of energy consumption baseline data corresponding to the selected building from the baseline data stores 704 and performs the functions described above to generate an energy lag for the building along with optimal model parameters. The energy lag is provided to the dispatch processor 702 on bus OPTLAG and the optimal model parameters are provided on bus OPTMPAR. The dispatch processor 702 may subsequently select a second building from the one or more buildings and provide its data to the lag optimizer 703 for generation of a second energy lag and second optimal model parameters. The dispatch processor 702 may subsequently perform these functions for remaining buildings in the participant stores 701. In one embodiment, the dispatch processor 702 and building lag optimizer 703 may perform the disclosed functions serially for each of the one or more buildings. In another embodiment, the dispatch processor 702 and lag optimizer 703 may perform the disclosed functions for a plurality of the one or more buildings concurrently.
In one embodiment, the dispatch processor 702 may employ the energy lags for corresponding ones of the one or more buildings to generate a dispatch schedule where buildings with greater energy lags are dispatched prior to buildings with lesser energy lags. The dispatch schedule is provided to the dispatch control 705 via DSCHED. Upon commencement of the program event, the dispatch control 705 controls the specified load shedding by performing load shedding actions in the order provided for by the dispatch schedule.
Referring to
The dispatch prediction element 802 is coupled to a building lag optimizer 803 via a second facility data bus FACDATA2, an optimal lag bus OPTLAG, and an optimal model parameters bus OPTMPAR. The building lag optimizer 803 is coupled to a baseline data stores 804.
The weather stores 806 comprises weather predictions that include outside temperatures corresponding to one or more buildings that are stored in the participant stores 801. The weather stores 806 may be located on site or may be located remotely and accessed via conventional networking technologies.
The participant stores 801 comprises identification data corresponding to one or more buildings that participate in the demand response program to allow for selection of corresponding baseline energy consumption data that is stored in the baseline data stores 804, to allow for energy lags and optimal regression model parameters to be associated therewith, and to further allow for employment of the energy lags and optimal regression model parameters, in conjunction with predicted outside temperatures provided via the weather stores 806, to estimate the first future time when the dispatch order is expected to be received for dispatch control of each of the participants in the program event. In one embodiment, the baseline data stores 804 comprises fine-grained baseline energy consumption data corresponding to the one or more buildings as described earlier, and functions in substantially the same manner as the baseline data stores 401 described above with reference to
Operationally, the demand response dispatch prediction system 800 is employed to estimate cumulative energy consumption as a function of the predicted outside temperatures occurring in a timeline for all of the one or more buildings that participate in the demand response program, where energy lags according to the present invention are utilized in generation of a cumulative energy consumption timeline. It is noted that, according to features of the present invention disclosed herein, the predicted energy consumption timeline may be employed to anticipate reception of a dispatch order to a finer level of granularity than that which has heretofore been provided. By using the energy lags associated with the buildings in the participant stores 801, estimated reception of a dispatch may be fine tuned. That is, using conventional dispatch prediction mechanisms that do not take into account energy lags of program participants may result in predicted dispatch reception times that are much earlier than necessary. Advantageously, by utilizing the present invention to determine a time when a dispatch threshold of energy consumption will be reached due to outside temperature, an energy services company or other demand response dispatch control entity may be provided with, say, additional hours for preparation of dispatch control actions.
The system 800 generates a predicted dispatch time that is provided to the dispatch control 805 for preparation of actions required to control each of the one or more buildings to optimally shed the energy specified in the dispatch order, upon reception of the dispatch order.
To predict the dispatch time, the dispatch prediction element 802 retrieves facility data from the participant stores 801 for a selected building and provides this data to the lag optimizer 803 on FACDATA2. Responsively, the lag optimizer 803 retrieves one or more sets of energy consumption baseline data corresponding to the selected building from the baseline data stores 804 and performs the functions described above to generate an energy lag for the building along with optimal model parameters. The energy lag is provided to the dispatch prediction element 802 on bus OPTLAG and the optimal model parameters are provided on bus OPTMPAR. The dispatch prediction element 802 subsequently selects a second building from the one or more buildings and provides its data to the lag optimizer 803 for generation of a second energy lag and second optimal model parameters. The dispatch prediction element 802 may subsequently perform these functions for remaining buildings in the participant stores 801. In one embodiment, the dispatch prediction element 802 and building lag optimizer 803 may perform the disclosed functions serially for each of the one or more buildings. In another embodiment, the dispatch prediction element 802 and lag optimizer 803 may perform the disclosed functions for a plurality of the one or more buildings concurrently.
Once the energy lags and optimal model parameters have been generated for all of the buildings in the participant stores 801, the dispatch prediction element 802 accesses the weather stores 806 to obtain future outside temperatures corresponding to each of the one or more buildings for a specified future time period. The dispatch predication element 802 then builds a cumulative future energy consumption timeline for all of the buildings using the outside temperatures as inputs to energy consumption models according to the present invention for all of the buildings. The dispatch prediction element 802 then processes the cumulative energy consumption timeline to determine a time when cumulative energy consumption increases as to cross a specified threshold known to trigger a demand response program event. The point at which consumption crosses the specified threshold is tagged as a dispatch time. From the dispatch time, the dispatch prediction element 802 may utilize demand response program contract data stored therein to calculate a predicted dispatch reception time, typically 24 hours prior to commencement of the demand response program event. The dispatch reception time is provided to the dispatch control element 805 on bus DISPTIME to allow for commencement of dispatch actions at a time having greater accuracy than that which has heretofore been provided.
Finally turning to
The peak prediction element 902 is coupled to a building lag optimizer 903 via a second facility data bus FACDATA2, an optimal lag bus OPTLAG, and an optimal model parameters bus OPTMPAR. The building lag optimizer 903 is coupled to a baseline data stores 904.
The weather stores 906 comprises weather predictions that include outside temperatures corresponding to buildings that are stored in the grid stores 901. The weather stores 906 may be located on site or may be located remotely and accessed via conventional networking technologies.
The grid stores 901 comprises identification data corresponding to buildings or aggregates of buildings that are part of the grid to allow for selection of corresponding baseline energy consumption data that is stored in the baseline data stores 904, to allow for energy lags and optimal regression model parameters to be associated therewith, and to further allow for employment of the energy lags and optimal regression model parameters, in conjunction with predicted outside temperatures provided via the weather stores 906, to estimate the future brown out time. For purposes of this discussion, aggregates of buildings may correspond to a unit of distribution over the grid such as, but not limited to, an electrical substation. As the present invention has been applied above to determine energy lags and optimal model parameters associated with single buildings, the present invention may also be applied to groups of buildings, say, a plurality of houses and business that are all powered from the same substation. Accordingly, the substation itself may be treated as a building for purposes of determining an energy lag and optimal model parameters. Henceforth, an aggregate of buildings will be simply referred to as a building.
In one embodiment, the baseline data stores 904 comprises fine-grained baseline energy consumption data corresponding to the buildings as described earlier, and functions in substantially the same manner as the baseline data stores 401 described above with reference to
Operationally, the brown out prediction system 900 is employed to estimate cumulative energy consumption on the grid as a function of the predicted outside temperatures occurring in a timeline for all of the buildings within the grid, where energy lags according to the present invention are utilized in generation of a cumulative energy consumption timeline. It is noted that, according to features of the present invention disclosed herein, the predicted energy consumption timeline may be employed to anticipate activation of the exceptional measures to a finer level of granularity than that which has heretofore been provided. By using the energy lags associated with the buildings in the grid stores 901, estimated time of occurrence of exceeding nominal production capacity may be fine tuned. That is, using conventional brown out prediction mechanisms that do not take into account energy lags of grid consumers may result in predicted brown out times that are much sooner than they actually occur. Advantageously, by utilizing the present invention to determine a time when a peak threshold of energy consumption will be reached due to outside temperature, a grid control entity may be provided with, say, additional hours to manage peak consumption on the grid.
The system 900 generates a predicted brown out time that is provided to the peak control 905 for preparation of exceptional measures required to manage peak consumption.
To predict the brown out time, the peak prediction element 902 retrieves building data from the grid stores 901 for a selected building and provides this data to the lag optimizer 903 on FACDATA2. Responsively, the lag optimizer 903 retrieves one or more sets of energy consumption baseline data corresponding to the selected building from the baseline data stores 904 and performs the functions described above to generate an energy lag for the building along with optimal model parameters. The energy lag is provided to the peak prediction element 902 on bus OPTLAG and the optimal model parameters are provided on bus OPTMPAR. The peak prediction element 902 subsequently selects a second building and provides its data to the lag optimizer 903 for generation of a second energy lag and second optimal model parameters. The peak prediction element 902 subsequently performs these functions for remaining buildings in the grid stores 901. In one embodiment, the peak prediction element 902 and building lag optimizer 903 may perform the disclosed functions serially for each of the one or more buildings. In another embodiment, the dispatch prediction element 902 and lag optimizer 903 may perform the disclosed functions for a plurality of the one or more buildings concurrently.
Once the energy lags and optimal model parameters have been generated for all of the buildings in the grid stores 901, the peak prediction element 902 accesses the weather stores 902 to obtain future outside temperatures corresponding to each of the one or more buildings for a specified future time period. The peak predication element 902 then builds a cumulative future energy consumption timeline for all of the buildings using the outside temperatures as inputs to energy consumption models according to the present invention for all of the buildings. The peak prediction element 902 then processes the cumulative energy consumption timeline to determine a time when cumulative energy consumption increases as to cross a specified threshold known to trigger the exceptional measures. The point at which consumption crosses the specified threshold is tagged as a brown out time. The brown out time is provided to the peak control element 905 on bus BOTIME to allow for commencement of the exceptional measures at a time having greater accuracy than that which has heretofore been provided.
Portions of the present invention and corresponding detailed description are presented in terms of software, or algorithms and symbolic representations of operations on data bits within a computer memory. These descriptions and representations are the ones by which those of ordinary skill in the art effectively convey the substance of their work to others of ordinary skill in the art. An algorithm, as the term is used here, and as it is used generally, is conceived to be a self-consistent sequence of steps leading to a desired result. The steps are those requiring physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of optical, electrical, or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
Throughout this disclosure, exemplary techniques and mechanisms have been employed in order to clearly teach features of the present invention. For instance, the thermal response processor discussed with reference to
It should be borne in mind, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless specifically stated otherwise, or as is apparent from the discussion, terms such as “processing” or “computing” or “calculating” or “determining” or “displaying” or the like, refer to the action and processes of a computer system, a microprocessor, a central processing unit, or similar electronic computing device, that manipulates and transforms data represented as physical, electronic quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system memories or registers or other such information storage, transmission or display devices.
Note also that the software implemented aspects of the invention are typically encoded on some form of program storage medium or implemented over some type of transmission medium. The program storage medium may be electronic (e.g., read only memory, flash read only memory, electrically programmable read only memory), random access memory magnetic (e.g., a floppy disk or a hard drive) or optical (e.g., a compact disk read only memory, or “CD ROM”), and may be read only or random access. Similarly, the transmission medium may be metal traces, twisted wire pairs, coaxial cable, optical fiber, or some other suitable transmission medium known to the art. The invention is not limited by these aspects of any given implementation.
The particular embodiments disclosed above are illustrative only, and those skilled in the art will appreciate that they can readily use the disclosed conception and specific embodiments as a basis for designing or modifying other structures for carrying out the same purposes of the present invention, and that various changes, substitutions and alterations can be made herein without departing from the scope of the invention as set forth by the appended claims.
This application is a continuation of U.S. patent application Ser. No. 14/674,033, filed on Mar. 31, 2015, having a common assignee and common inventors, and which is herein incorporated by reference for all intents and purposes. This application is related to the following co-pending U.S. Patent Applications, each of which has a common assignee and common inventors. FILINGSER. NO.DATETITLE14/674,004Mar. 31, 2015SYSTEM FOR WEATHER INDUCED FACILITY ENERGYCONSUMPTION CHARACTERIZATION14/984,612Dec. 30, 2015WEATHER INDUCED FACILITY ENERGYCONSUMPTION CHARACTERIZATION MECHANISM14/674,021Mar. 31, 2015DEMAND RESPONSE DISPATCH SYSTEMEMPLOYING WEATHER INDUCED FACILITY ENERGYCONSUMPTION CHARACTERIZATIONS14/984,706Dec. 30, 2015APPARATUS AND METHOD FOR EMPLOYINGWEATHER INDUCED FACILITY ENERGYCONSUMPTION CHARACTERIZATIONS IN ADEMAND RESPONSE DISPATCH SYSTEM15/903,533Feb. 23, 2018DISPATCH SYSTEM EMPLOYING WEATHERINDUCED FACILITY ENERGY CONSUMPTIONCHARACTERIZATIONS15/903,596Feb. 23, 2018SYSTEM FOR DEMAND RESPONSE DISPATCHEMPLOYING WEATHER INDUCED FACILITY ENERGYCONSUMPTION CHARACTERIZATIONS15/903,651Feb. 23, 2018WEATHER INDUCED DEMAND RESPONSEDISPATCH SYSTEM15/903,705Feb. 23, 2018DEMAND RESPONSE DISPATCH SYSTEMEMPLOYING WEATHER INDUCED ENERGYCONSUMPTION14/984,785Dec. 30, 2015APPARATUS AND METHOD FOR PREDICTION OF ADEMAND RESPONSE DISPATCH15/961,073Apr. 24, 2018SYSTEM FOR DEMAND RESPONSE DISPATCHPREDICTION EMPLOYING WEATHER INDUCEDFACILITY ENERGY CONSUMPTIONCHARACTERIZATIONS14/674,041Mar. 31, 2015ENERGY BROWN OUT PREDICTION SYSTEM14/674,041Dec. 30, 2015APPARATUS AND METHOD FOR PREDICTION OF ANENERGY BROWN OUT14/674,057Mar. 31, 2015APPARATUS AND METHOD FOR DEMANDCOORDINATION NETWORK CONTROL
| Number | Name | Date | Kind |
|---|---|---|---|
| 4090088 | McMahon et al. | May 1978 | A |
| 4168491 | Phillips et al. | Sep 1979 | A |
| 4228364 | Walden | Oct 1980 | A |
| 7324876 | Ying | Jan 2008 | B2 |
| 7552100 | Chen | Jun 2009 | B2 |
| 7565227 | Richard et al. | Jul 2009 | B2 |
| 7881889 | Barclay et al. | Feb 2011 | B2 |
| 8019567 | Steinberg et al. | Sep 2011 | B2 |
| 8355827 | Egnor et al. | Jan 2013 | B2 |
| 8370283 | Pitcher et al. | Feb 2013 | B2 |
| 8401709 | Cherian et al. | Mar 2013 | B2 |
| 8417391 | Rombouts et al. | Apr 2013 | B1 |
| 8706650 | Ozog | Apr 2014 | B2 |
| 8751054 | Wang et al. | Jun 2014 | B2 |
| 9153965 | Deshpande et al. | Oct 2015 | B2 |
| 9158322 | Willig | Oct 2015 | B2 |
| 9196009 | Drees et al. | Nov 2015 | B2 |
| 9244444 | Carty et al. | Jan 2016 | B2 |
| 9292888 | An et al. | Mar 2016 | B2 |
| 9355069 | ElBsat | May 2016 | B2 |
| 9569804 | Stein et al. | Feb 2017 | B2 |
| 20040254686 | Matsui et al. | Dec 2004 | A1 |
| 20050043862 | Brickfield et al. | Feb 2005 | A1 |
| 20090012653 | Cheng | Jan 2009 | A1 |
| 20090177520 | Bateni et al. | Jul 2009 | A1 |
| 20100222935 | Forbes, Jr. et al. | Sep 2010 | A1 |
| 20100286937 | Hedley et al. | Nov 2010 | A1 |
| 20100318227 | Steinberg et al. | Dec 2010 | A1 |
| 20110119042 | Johnson et al. | May 2011 | A1 |
| 20110130886 | Drees et al. | Jun 2011 | A1 |
| 20110190951 | Lee | Aug 2011 | A1 |
| 20110231320 | Irving | Sep 2011 | A1 |
| 20120046796 | Zhang | Feb 2012 | A1 |
| 20120185106 | Ghosh et al. | Jul 2012 | A1 |
| 20120278051 | Jiang et al. | Nov 2012 | A1 |
| 20120330626 | An et al. | Dec 2012 | A1 |
| 20130024141 | Marwah et al. | Jan 2013 | A1 |
| 20130041853 | Shnekendorf et al. | Feb 2013 | A1 |
| 20130073273 | Madrazo et al. | Mar 2013 | A1 |
| 20130297092 | Willig et al. | Nov 2013 | A1 |
| 20130304269 | Shiel | Nov 2013 | A1 |
| 20130346768 | Forbes, Jr. et al. | Nov 2013 | A1 |
| 20140148953 | Nwankpa et al. | May 2014 | A1 |
| 20140297238 | Parthasarathy et al. | Oct 2014 | A1 |
| 20150006294 | Irmak et al. | Jan 2015 | A1 |
| 20150057820 | Kefayati et al. | Feb 2015 | A1 |
| 20150076926 | Kosaka et al. | Mar 2015 | A1 |
| 20150088442 | Farrar et al. | Mar 2015 | A1 |
| 20150169795 | ElBsat et al. | Jun 2015 | A1 |
| 20150198961 | Shiel | Jul 2015 | A1 |
| 20150198962 | Shiel | Jul 2015 | A1 |
| 20150207752 | Birkestrand et al. | Jul 2015 | A1 |
| 20150227846 | Mercer et al. | Aug 2015 | A1 |
| 20160195887 | Shiel | Jul 2016 | A1 |
| 20160266181 | Kawaguchi et al. | Sep 2016 | A1 |
| 20160281607 | Asati et al. | Sep 2016 | A1 |
| 20160291067 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160291561 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160291562 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160291616 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160291622 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160291623 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160291624 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160292577 | Al-Mohssen et al. | Oct 2016 | A1 |
| 20160294185 | Al-Mohssen et al. | Oct 2016 | A1 |
| 20160294186 | Ai-Mohssen et al. | Oct 2016 | A1 |
| 20160305678 | Pavlovski et al. | Oct 2016 | A1 |
| 20160363948 | Steven et al. | Dec 2016 | A1 |
| 20180341283 | Ai-Mohssen et al. | Nov 2018 | A1 |
| 20180356849 | Ai-Mohssen et al. | Dec 2018 | A1 |
| 20180356850 | Ai-Mohssen et al. | Dec 2018 | A1 |
| 20180356851 | Ai-Mohssen et al. | Dec 2018 | A1 |
| 20190004555 | Al-Mohssen et al. | Jan 2019 | A1 |
| 20190018438 | Al-Mohssen et al. | Jan 2019 | A1 |
| 20190044328 | Al-Mohssen et al. | Feb 2019 | A1 |
| 20190044329 | Al-Mohssen et al. | Feb 2019 | A1 |
| 20190044330 | Al-Mohssen et al. | Feb 2019 | A1 |
| 20190044331 | Al-Mohssen et al. | Feb 2019 | A1 |
| Entry |
|---|
| Conca, James. “Demand Response Is How the Smart Grid Will Save Us Billions.” Feb. 24, 2015 Forbes. pp. 1-5. |
| Lamonica, Martin. “Green Biz 101: What do you Need to Know About Demand Response?” Apr. 29, 2014 GreenBiz 101. pp. 1-9. |
| Number | Date | Country | |
|---|---|---|---|
| 20180239382 A1 | Aug 2018 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 14674033 | Mar 2015 | US |
| Child | 15961020 | US |
