U.S. patent application number 13/493429 was filed with the patent office on 2013-12-12 for high efficiency water heater.
The applicant listed for this patent is George R. Arnold. Invention is credited to George R. Arnold.
Application Number | 20130327313 13/493429 |
Document ID | / |
Family ID | 49714302 |
Filed Date | 2013-12-12 |
United States Patent
Application |
20130327313 |
Kind Code |
A1 |
Arnold; George R. |
December 12, 2013 |
HIGH EFFICIENCY WATER HEATER
Abstract
A high-efficiency water heating system includes at least one
source of heat and a processor interfaced to the at least one
source of heat. The processor controls the operation of the at
least one source of heat (e.g. energizes a heating element to
provide heat to the water). At least one source of data related to
a consumption of hot water from the high-efficiency water heating
system is provided. Software running on the processor analyzes the
data and calculates a predicted demand for the hot water based upon
the data, then controls the operation of the at least one source of
heat in response to the predicted demand.
Inventors: |
Arnold; George R.;
(Sutherland, FL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Arnold; George R. |
Sutherland |
FL |
US |
|
|
Family ID: |
49714302 |
Appl. No.: |
13/493429 |
Filed: |
June 11, 2012 |
Current U.S.
Class: |
126/344 |
Current CPC
Class: |
F24H 9/2021 20130101;
F24H 9/2007 20130101 |
Class at
Publication: |
126/344 |
International
Class: |
F24H 9/20 20060101
F24H009/20; F24H 1/18 20060101 F24H001/18 |
Claims
1. A high-efficiency water heating system comprising: at least one
source of heat, the source of heat interfaced to a water supply
such that the source of heat controllably transfers heat into water
from the water supply, thereby producing heated water; a processor
interfaced to the source of heat, the processor controlling the
operation of the source of heat; at least one source of data, the
data related to a consumption of the heated water from the
high-efficiency water heating system; and software running on the
processor, the software analyzing the data and the software
calculating a predicted demand for the heated water and the
software controlling the operation of the at least one source of
heat responsive to the predicted demand.
2. The high-efficiency water heating system of claim 1, wherein the
at least one source of data is selected from the group consisting
of indoor ambient temperature, indoor ambient humidity, outdoor
ambient temperature, outdoor ambient humidity.
3. The high-efficiency water heating system of claim 1, further
comprising a network interface, the network interface operatively
coupled to the processor and the network interface providing a
connection between the processor and a network.
4. The high-efficiency water heating system of claim 3, wherein the
at least one source of data includes external data from an external
provider, the external provider connected to the processor through
the network and the processor accessing the external data through
the network and the network interface.
5. The high-efficiency water heating system of claim 1, further
comprising a water storage tank, at least one of the at least one
sources of heat being operatively coupled to the water storage tank
such that, when the at least one sources of heat that is
operatively coupled to the water storage tank is energized, heat is
transferred to water within the storage tank to produce the heated
water.
6. The high-efficiency water heating system of claim 5, wherein at
least one of the at least one sources of heat is an on-demand
heating element.
7. The high-efficiency water heating system of claim 6, further
comprising at least one electrically controlled valve, each of the
at least one electrically controlled valve is operatively connected
to the processor such that, under control of the processor, heated
water is supplied from the water storage tank, the on-demand
heating element(s), or a combination of both the water storage tank
and the on-demand heating element(s).
8. The high-efficiency water heating system of claim 1, further
comprising a flow sensor, the flow sensor operatively coupled to
the processor, and the flow sensor providing to the processor a
measurement of current demand for the heated water.
9. The high-efficiency water heating system of claim 1, wherein the
software running on the processor uses neural networks to calculate
the predicted demand.
10. A method of providing heated water, the method comprising: (a)
calculating a predicted demand for heated water over a period of
time; (b) prior to and during the period of time, signaling one or
more sources of heat to energize based upon the predicted demand,
the sources of heat interfaced to a water supply such that the
sources of heat provide heat to water to produce a quantity of the
heated water; (c) measuring a rate of flow of the heated water
during the period of time; and (d) feeding the rate of flow of the
heated water back into the step of calculating to improve the
accuracy of the step of calculating; and (e) repeating steps
a-e.
11. The method of claim 10, wherein the step of calculating uses a
neural network.
12. The method of claim 10, wherein the step of calculating takes
into account ambient conditions.
13. The method of claim 10, wherein the step of calculating takes
into account a schedule of users of the heated water.
14. The method of claim 10, wherein the step of calculating takes
into account external data.
15. The method of claim 10, wherein the step of calculating takes
into account historic demand for heated water.
16. A high-efficiency water heating system comprising: at least
source of heat operatively coupled to a water storage tank such
that, when the at least one source of heat is energized, heat is
transferred into water within the storage tank, thereby producing
heated water; at least one on-demand heating device connected to a
supply of water such that, when the at least one on-demand heating
device is energized, heat is transferred into water from a supply
of water by the at least one on-demand heating device, thereby
producing the heated water; a first electrically controlled valve
connected between the output of the storage tank and heated water
plumbing; a second electrically controlled valve connected between
the output of the at least one on-demand heating device and the
heated water plumbing; a processor interfaced to the at least one
source of heat and to the at least one on-demand heating device,
the processor controlling the operation of the at least one source
of heat, the at least one on-demand heating device, the first
valve, and the second valve; a flow sensor, the flow sensor
operatively coupled to the processor, and the flow sensor providing
a measurement of current demand for the heated water; at least one
source of data, the data related to a consumption of the heated
water from the high-efficiency water heating system; and software
running on the processor, the software analyzing the data and the
measurement of current demand, the software calculating a predicted
demand for the heated water, and the software controlling the
operation of the at least one source of heat, the at least one
on-demand heating device, the first valve, and the second valve,
responsive to the predicted demand.
17. The high-efficiency water heating system of claim 16, wherein
the at least one source of data includes measurement data selected
from the group consisting of indoor ambient temperature, indoor
ambient humidity, outdoor ambient temperature, outdoor ambient
humidity.
18. The high-efficiency water heating system of claim 16, wherein
the software further records historical data that includes data
from the flow sensor such that, in the future, the software uses
the historical data to adjust the predicted demand based upon both
the data and the historical data.
19. The high-efficiency water heating system of claim 16, wherein
the software running on the processor uses neural networks to
calculate the predicted demand.
20. The high-efficiency water heating system of claim 16, further
comprising a plurality of sensors, the sensors interfaced to the
processor, the sensors providing at least part of the data, wherein
the sensors are selected from the group consisting of temperature
sensors, humidity sensors, and ambient light sensors.
Description
FIELD
[0001] This invention relates to the field of providing hot water
and more particularly to a system, apparatus, and method for
predictively heating hot water.
BACKGROUND
[0002] It has been estimated that water heaters account for up to
30 percent of an average homes energy budget. Currently, there are
two main classifications of home and industrial water heaters. The
first main classification of water heater is known as a "convention
storage tank" water heater. This type of water heater is commonly
used in many homes and businesses and utilizes a source of heat and
a storage tank. Heat from a source such as electrical heating
elements or fossil fuel burners increases the temperature of water
in the storage tank until it reaches a pre-determined temperature,
at which time the heat source is shut off (conserving energy) until
the temperature of the water in the storage tank drops to a second
pre-determined temperature. Such systems provide for large,
transient demands for hot water by providing large storage tanks or
heating the water in the storage tanks to a high temperature, then
mixing the hot water with unheated water before distributing the
water to the end users. Although the storage tanks are typically
thermally insulated, these systems lose efficiency due to heat loss
due to conduction from the storage tanks to the ambient environment
and through the plumbing that connects the storage tanks to the
water supply and delivery plumbing, especially in situations where
there are extended periods of time during which no hot water is
used. Such situations occur, for example, in a residence when
household members are at work or sleeping, or in an industrial
facility when all of the workers are home.
[0003] The second main classification of water heater is known as
an "on-demand" water heater. This type of water heater has no
storage tank, instead having a very high energy output heating
element that is capable of raising the temperature of the water
from ground water temperature to the desired hot water temperature
as the hot water is used. Typically, the on-demand water heater has
a flow sensor and a heating element such that, as soon as demand
for hot water occurs, the sensor detects that water is flowing
through the hot water plumbing and enables the heating element,
which heats the water as the water flows through the heating
element. This class of water heaters is more efficient than
storage-tank based water heaters because there is little or no
energy loss when no hot water is being used. This class of water
heaters has its own set of drawbacks. These water heaters require a
large amount of energy during operation to enable the quick heating
of water from ground water temperatures to the desired hot water
temperature at a given flow rate. As a result of this, such
on-demand water heaters are often rated for a certain rise in
temperature for a given flow rate. For example, one on-demand water
heater is capable of a 45 F rise in temperature at a flow rate of
8.5 gallons per minute, but only a 35 F rise in temperature at a
9.5 gallon per minute rate. Therefore, if several people are
concurrently taking a shower, the hot water will not be as hot.
Furthermore, because there is no storage tank containing pre-heated
water, if there is a power failure, there will be no hot water
until power is restored. Another setback of on-demand water heaters
is supplying sufficient power for the heating elements. Natural gas
is often used because it provides a high amount of BTUs, but not
all homes have natural gas. Electricity is more widely available,
but most existing buildings do not have sufficient power and/or
wiring for whole-house on-demand water heaters.
[0004] The storage-tank based hot water system has many advantages
such as supplying sufficient hot water for most anticipated
demands, constant hot water temperatures, providing a fair amount
of hot water during a power outage, etc. Therefore, storage-tank
based hot water systems will continue to be used.
[0005] Prior attempts at making improvements involved simple timers
that were manually set to prevent water heating during periods of
no demand (e.g. when families are sleeping), systems that
understand variable rate plans (e.g. electricity costs more during
peak periods) to adjust heating patterns to the electricity costs,
and systems that are remotely controlled by the electric companies
to reduce electricity consumption during high demand periods,
therefore preventing a grid overload.
[0006] What is needed is a system that will improve the overall
efficiency of the storage-tank based hot water heater systems by
predicting future usage patterns.
SUMMARY
[0007] In one embodiment, a high-efficiency water heating system is
disclosed including at least one source of heat and a processor
interfaced to the at least one source of heat, The processor
controls the operation of the at least one source of heat (e.g.
energizes the at least one heating element to provide heat to the
water). At least one source of data related to a consumption of hot
water from the high-efficiency water heating system is provided and
software running on the processor analyzes the data and calculates
a predicted demand for the hot water based upon the data, then
controls the operation of the at least one source of heat
responsive to the predicted demand.
[0008] In another embodiment, a method of providing hot water is
disclosed including (a) calculating a predicted demand for hot
water over a period of time, then, (b) prior to and during the
period of time, energizing one or more sources of heat based upon
the predicted demand, the sources of heat interfaced to a water
supply such that the source of heat provide heat to water from the
water supply and produce a quantity of the heated water. (c) A rate
of flow of the heated water during the period of time is measured;
and (d) the rate of flow of the heated water is fed back into the
step of calculating to improve the accuracy of the step of
calculating. Steps a-e are repeated.
[0009] In another embodiment, a high-efficiency water heating
system is disclosed including at least one source of heat
operatively coupled to a water storage tank. When the at least one
source of heat is energized, heat is transferred into water within
the storage tank. At least one on-demand heating device (e.g. gas
burner or electric element) is connected to a supply of water such
that, when the at least one on-demand heating element is energized,
heat is transferred into water from the supply of water by the at
least one on-demand heating device. Two valves control the flow of
water: a first electrically controlled valve connecting the output
of the storage tank to supply the heated water and a second
electrically controlled valve connecting the output of the at least
one on-demand heating device to supply the heated water. A
processor is interfaced to the at least one source of heat, to the
at least one on-demand heating device, to the first valve, and to
the second valve, the processor controlling the operation of the at
least one source of heat, the at least one on-demand heating
device, the first valve, and the second valve. A flow sensor is
operatively coupled to the processor, providing a measurement of
current demand for the heated water. At least one source of data
related to a consumption of the hot water from the high-efficiency
water heating system is provided. Software running on the processor
analyzes the data and calculates a predicted demand for the hot
water and controls the operation of the at least one source of
heat, the at least one on-demand heating device, the first valve,
and the second valve, responsive to the predicted demand.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The invention can be best understood by those having
ordinary skill in the art by reference to the following detailed
description when considered in conjunction with the accompanying
drawings in which:
[0011] FIG. 1 illustrates a schematic view of a high-efficiency hot
water heater software system.
[0012] FIG. 2 illustrates a schematic view of the high-efficiency
hot water heater decision process.
[0013] FIG. 3 illustrates a schematic view of a typical computer
system of the high-efficiency hot water heater.
[0014] FIG. 4A illustrates a schematic view of a typical
storage-tank hot water heater of the prior art.
[0015] FIG. 4B illustrates a schematic view of a typical on-demand
hot water heater of the prior art.
[0016] FIG. 5 illustrates a simplified schematic view of the
high-efficiency hot water heater.
[0017] FIG. 6 illustrates a second schematic view of the
high-efficiency hot water heater.
[0018] FIG. 7 illustrates a flow chart of a storage-tank hot water
heater of the prior art.
[0019] FIG. 8 illustrates a flow chart of an on-demand hot water
heater of the prior art.
[0020] FIG. 9 illustrates a flow chart of a high-efficiency hot
water heater.
[0021] FIG. 10 illustrates a schematic diagram of a
neural-network-based high-efficiency hot water heater.
DETAILED DESCRIPTION
[0022] Reference will now be made in detail to the presently
preferred embodiments of the invention, examples of which are
illustrated in the accompanying drawings. Throughout the following
detailed description, the same reference numerals refer to the same
elements in all figures.
[0023] Throughout this description, the term "heating element"
refers to any type of heating element, including, but not limited
to, electric heating elements, gas burners, oil burners, etc. The
heating element(s), when energized, provide heat to a body of
water. Note that the colloquial meaning of water includes H.sub.20
and any impurities present in a water supply such as other
chemicals, minerals, solids, and dissolved gasses.
[0024] Throughout the description, the terms "hot water" and
"heated water" refers to water that has been heated above the
ambient water temperature as one would expect to find when opening
a left spigot on a sink or one that is marked `H` in English
speaking countries like the United States (note, other marks exist
for other countries such as `C` for Caldo in Italy).
[0025] Throughout the description, embodiments are shown having a
certain number and type/location of storage tanks, heating elements
and valves. The high-efficiency hot water heater is not restricted
in any way to any particular number of storage tanks, heating
elements or heat sources, and/or valves. For example, in some
embodiments, there are no valves and in some embodiments, there is
only one "mixing" valve. The high-efficiency hot water heater is
not restricted in any way to any specific source of heat
configuration/combination. For example, some storage tank systems
have upper and lower heating elements while others have only a
single gas burner, while still other systems have heat sources
(e.g. boilers, solar panels), that are external to the storage
tanks. There are no restrictions as to the types of heat sources
and/or locations of the heat sources.
[0026] Although described as a processor-based system, as known,
any processor-based system is capable of being made using discrete
components (e.g. gates) and such implementations are anticipated
and included here within.
[0027] Any form of control of the valves and heating elements is
anticipated, including, but not limited to, electrical control
through individual wiring, electrical control through a wired
network, electrical control through a wireless system, pneumatic
control, etc.
[0028] Throughout this description, various examples of home and/or
industrial water heating scenarios are provided. The disclosed and
claimed invention is not limited in any way to a particular
application and is intended for use in any water heating
application.
[0029] Referring to FIG. 1, a schematic view of a high-efficiency
hot water heater software system is shown. The system shown in FIG.
1 applies to some embodiments of the high-efficiency hot water
heater software system. Although the high-efficiency hot water
heater software system is intended to operate with or without a
network 10, it is shown interfaced to the Internet 10 (a.k.a. the
World Wide Web).
[0030] In this embodiment, one or more heater controllers 20 are
connected to the network 10 as known in the industry. One or more
servers 40 are also connected to the network 10 as known in the
industry. The high-efficiency hot water heater software system
includes data 42 for authentication as well as history, etc. It is
anticipated that any or all of the storage areas 42 are locally
interfaced to the server 40, remotely interfaced to the server 40
(e.g., Network Attached Storage--NAS) and/or remotely interfaced to
the server 40 over a network, either a local area network or wide
area network.
[0031] In some embodiments, the server 40 or the individual heater
controllers 20 also interfaces to various providers of external
data 50. The information providers 50 are interfaced to the server
40 and/or heater controllers 20 with any known network or direct
connection, as known in the industry. As shown in the example of
FIG. 1, the information providers 50 are interfaced to the server
40 through the Internet 10. When used, the information providers 50
deliver external data to the server 40 and/or heater controllers 20
related to hot water demand such as weather predictions, sunrise
and sunset times, etc. The high-efficiency hot water heater uses
this data, when available, to help predict future hot water demand.
For example, if it will be cloudy, more hot water is likely to be
needed because people taking showers feel colder on cloudy
days.
[0032] The architecture shown in FIG. 1 is flexible, in that, the
decision process is made either locally at the hot water heater
controllers 20 (see FIGS. 3, 5 and 6) or centrally at a central
computing system (server) 40 or jointly. Therefore, several
high-efficiency hot water heater system
architectures/configurations are anticipated. In some
high-efficiency hot water heater systems, a controller 20 (see FIG.
3) is directly interfaced to the heating elements 120/122 (see
FIGS. 5 and 6), obtaining data locally and/or remotely (e.g. from
Information Providers 50) and making all decisions on when and how
to heat water. In some high-efficiency hot water heater systems,
the controller 20 or other discrete logic is interfaced to the
heating elements 120/122 (see FIGS. 5 and 6) and is controlled by
software running on a remote system (e.g. server 40). In this, the
remote system (server 40) optionally receives data from the local
controller 20 and/or remotely (e.g. from Information Providers 50)
and makes all decisions on when and how to heat water, instructing
the controller 20 to turn on/off the heating elements 120/122. In
another type of high-efficiency hot water heater system, a
combination of the prior architectures is used. In this, the
controller 20 (see FIG. 3) or other discrete logic is directly
interfaced to the heating elements 120/122 (see FIGS. 5 and 6) and
makes part of the decisions as to when and how to heat the water,
but the processor 70 is also in communication and cooperating with
software running on a remote system (e.g. server 40). In this, the
remote system (server 40) optionally receives data from the local
controller 20 and/or remotely (e.g. from Information Providers 50)
and jointly, with the controller 20, makes decisions on when and
how to heat water.
[0033] Referring to FIG. 2, a schematic overview of the
high-efficiency hot water heater decision process is shown. In this
example, one or more inputs 32/33/34/35/36 are analyzed by analysis
algorithms 30 to determine how to control the high-efficiency hot
water heater heating elements 120/122 and or valves 130/132/134
(see FIGS. 5 and 6). Any or all of the inputs 32/33/34/35/36 are
used to predict when different amounts of hot water will be needed.
For example, ambient conditions 32 are used by the analysis
algorithms 30 to predict hot water needs. For example, a person
will adjust the water temperature of their shower dependent upon
internal building ambient temperature and humidity. A higher
internal building ambient temperature and humidity tends to make a
person lower their shower temperature while a lower internal
building ambient temperature and humidity tends to make a person
increase their shower temperature. Likewise, external ambient
conditions tend to further influence a person's shower temperature.
For example, when it is cloudy out, a person tends to take warmer
showers and when it is hot outside, after exercise outdoors, a
person tends to take cooler showers. Ambient conditions 32 are
measured by, for example, temperature sensors 100/104, humidity
sensors 110, and light sensors 112 (see FIG. 3).
[0034] Schedule data 33 is also optionally considered by the
analysis algorithms 30 to predict when demand will occur. For
example, in a dormitory, by knowing the schedule of students, the
analysis algorithms 30 control the heater(s) 120/122 and valves
130/132/134 based upon the schedule data 33 such that, knowing that
lights out starts at 10:00 PM and classes start at 8:00 AM, the
analysis algorithms 30 predict very low hot water usage after 10:00
PM when the students are asleep, high hot water usage prior to 8:00
AM when students are waking and taking showers, and low hot water
usage when students are in class after 8:00 AM, etc. In another
example, by knowing the schedule data 33 for people in a home, the
analysis algorithms 30 make similar predictions. For example, in a
two person household, knowing that both occupants leave work at
7:30 AM and return home at 6:00 PM on Monday, Tuesday and Friday,
the analysis algorithms 30 predict low or no hot water usage
starting at 7:30 AM and pre-heat the hot water in the storage tank
140 (see FIG. 5) to reach a certain temperature at 6:00 PM when the
occupants arrive home (e.g. for dish washing, showers, etc.). In
some examples, schedule data includes resource levels such as:
there are 40 residents in the dormitory, twenty start classes at
8:00 AM and 40 residents start classes at 9:15 AM. In this example,
the number of students is also used to determine hot water demand
(e.g. if there were more than 40 students, more hot water is
needed).
[0035] Current data 35 is also (optionally) considered by the
analysis algorithms 30 to predict when demand will occur. Current
data includes, for example, current flow rates, current hot water
output temperature, and current system input water temperature. The
analysis algorithms 30 use the current data 35 to determine how
well predictions match actual and if any boosting or setback is
needed based upon this current data 35. For example, if the
analysis algorithms 30 predict an overall demand of 600 gallons of
hot water per hour (10 gallons per minute) and the current data 35
indicates that in the last 5 minutes, 30 gallons per minute have
been consumed, then the analysis algorithms will control the water
heating elements 120/122 to energize and heat the water temperature
to a higher temperature than was predicted to be required.
Likewise, if for some reason the temperature of the input water is
a few degrees less than it has been; more heat needs to be added to
the system to account for colder water being mixed with the heated
water in the storage tank 140 (see FIGS. 5 and 6).
[0036] Another optional source of data to the analysis algorithms
30 is external data 36. External data includes any data feed that
has information regarding the future demand for hot water. This
data includes, but is not limited to, weather predictions, data
from an almanac (e.g. sunrise and sunset times), local news
information, school information, lunch menus, school events, local
events, etc. For example, if the dormitory dinner menu includes
cold sandwiches, there is likely to be a higher demand for hot
water that evening then if the menu includes hot soup. If the
forecast for tomorrow is rain and sleet, depending on the users,
such weather will change demand. For example, some hot water users
will forego showering/bathing until they return from classes so as
to not be as cold walking to classes. Other events will affect hot
water consumption. For example, if Friday night is Prom Night, then
extra hot water will be needed between, say, 5:00 PM and 7:00 PM
for prom goers to shower/bathe.
[0037] Although similar to schedule data 33, historic data 34
differs, in that demand measurements from the past are used to
predict future demand by the analysis algorithms 30. For example,
after installation of the high-efficiency hot water heater system,
the analysis algorithms 30 initially relies on ambient data 32,
schedule data 33, external data 36 and current data 35 to predict
hot water demand and control the hot water heaters 60. After a
number of days of controlling the high-efficiency hot water heater
based upon these data, the analysis algorithms now have access to
historical data 34. For example, one set of 40 students may have
hot water usage patterns that differ from another set of the same
number of students, perhaps due to differences in gender and other
backgrounds. Initially, the analysis algorithms 30 make an
assumption of the hot water needs per person, for example, making
an average consumption prediction or a worst-case consumption
prediction. As time goes on and usage patterns start repeating, the
analysis algorithms 30 consult the historic data 34 to determine
how the operation of the high-efficiency hot water heater needs to
be adjusted based upon the historic data 34. For example, the first
Monday after installation, the high-efficiency hot water heating
system preheats the water in storage to a certain temperature based
upon 40 unknown people. By measuring the flow rates, input water
temperature and output water temperature, the analysis algorithms
determine that the water in the storage tank 140 does not need to
be heated as much as it was, so the next Monday, the water in the
storage tank 140 is heated a percentage less (e.g. 10% less) and
the flow rates, input water temperature and output water
temperature are again measured to determine how well the
high-efficiency hot water heating system is meeting the demands of
users.
[0038] Although not required, it is anticipated that the analysis
algorithms 30 use neural networks 430 (see FIG. 10) or some other
form of artificial intelligence to analyze the data 32/33/34/35/36
and make appropriate decisions on when and how much to pre-heat the
water based upon predictions of such analysis algorithms. A neural
network 430 accepts data (as described above) and determines
actions (as described above), with the added benefit that the
neural network 430 learns. The neural networks 430 find patterns in
the data as well as filter the data. For example, over time, the
neural network 430 will find that every Friday morning between 6:00
and 6:15, there is a high demand for hot water and, therefore, will
make assumptions on how the water heater needs to be controlled
based upon the predicted demand. The neural network 430 will filter
its data to ignore irregular data. For example, when users are on
vacation and during one week, there is no demand for hot water on
that day.
[0039] In the neural network 430 implementation, each input is
considered with a weighing factor. For example, last week's usage
history has a high weighing factor, the week before usage history
has a lower weighing factor, and the external weather (e.g. cloudy,
raining) has even a lower weighing factor. As the neural network
system 430 continues to predict hot water demand, hot water usage
(e.g. flow rates) is measured and fed back into the neural network
430 and the neural network 430 makes adjustments. For example, if,
over time, the neural network 430 recognizes that hot water demand
is 10% higher on cloudy days, the neural network 430 will increase
the weight given to external weather.
[0040] Being that neural networks 430 are well known, it is
anticipated that for some high-efficiency hot water heaters, the
basic neural network 430 software is provided as a package from a
provider of such and is programmed based upon the range of inputs
available to the high-efficiency hot water heater (e.g., schedule,
ambient/weather, history, current data, external data, etc.) to
control the available heating elements 120/122 (see FIGS. 5 and 6)
and/or valves 130/132/134 (see FIGS. 6 and 7). For some
high-efficiency hot water heater systems, the neural network
software is not an off-the-shelf pre-programmed package.
[0041] In some high-efficiency hot water heaters, instead of using
true neural networks, heuristic algorithms or static logic is used
in the prediction algorithms 30. A simple example in a dormitory in
which all students leave for breakfast and class at the same time
and there are n students registered for that dormitory, an
exemplary heuristic algorithm is: if n is less than 30, preheat the
water to t1 at time T1; if n is greater than 30 and less than 60,
preheat the water to t2 at time T2; and if n is greater than 60,
preheat the water to t3 at time T3. In this, the schedule data 33
is used to determine when the students will be using the hot water
(e.g. before class/breakfast) and how many students are present.
The more students present, the hotter the water in the storage tank
needs to be, therefore, heating starts earlier and ends when the
water reaches this higher temperature. This is but an example and a
complete heuristic algorithm will consider other data in the
algorithm's decision tree.
[0042] Referring to FIG. 3, a schematic view of a typical computer
system is shown. The example computer system represents a typical
computer system used as the individual heater control devices 20,
though a similar computer system is anticipated for the server 40.
The exemplary computer system 20 is shown in its simplest form,
having a single processor 70. Many different computer architectures
are known that accomplish similar results in a similar fashion and
the present invention is not limited in any way to any particular
computer system. The present invention works well utilizing a
single processor system, as shown in FIG. 3, a multiple processor
system where multiple processors share resources such as memory and
storage, a multiple server system where several independent servers
operate in parallel (perhaps having shared access to the data or
any combination). In any of these systems, a processor 70 executes
or runs stored programs that are generally stored for execution
within a memory 74. The processor 70 is any processor or a group of
processors, for example an Intel Pentium-4.RTM. CPU, 80C51, or the
like. The memory 74 is connected to the processor by a memory bus
72 and is any memory 74 suitable for connection with the selected
processor 210, such as SRAM, DRAM, SDRAM, RDRAM, DDR, DDR-2, etc.
It is also anticipated that the processor 70, bus 72 and memory 74
are integrated into a single component.
[0043] Also connected to the processor 70 is a system bus 82 for
connecting to peripheral subsystems such as a network interface 80,
persistent storage (e.g. a hard disk, flash memory) 88, removable
storage (e.g. DVD, CD, flash drive) 90, a graphics adapter 84 and a
keyboard/mouse 92. The graphics adapter 84 receives commands and
display information from the system bus 82 and generates a display
image that is displayed on the display 86 (e.g. monitor, LEDs,
graphic display, etc.).
[0044] Various input devices, sensors, and control drivers
100/104/110/112/114/116/118 are optionally connected to the bus.
The following inputs are representative of inputs for the
high-efficiency hot water system, though more or less inputs are
anticipated: one or more outside ambient temperature sensors
(t.sub.a) 100, one or more indoor building temperature sensors
(t.sub.b) 104, one or more relative humidity sensors 110, one or
more outdoor ambient light sensors (e.g. cloud cover) 112, cold
water supply temperature sensor (t.sub.i) 115, hot water output
temperature sensor (t.sub.o) 117, and a water flow sensor 114.
[0045] The exemplary control outputs include one or more heater
controls 118 and one or more valve controls 116. The heater
controls 118 energize one or more water heating elements 120/122
(see FIGS. 5 and 6). The valve controls 116 energize (open or close
or partially open) one or more water control valves 130/132/134
(see FIGS. 5 and 6). Although shown as discrete inputs
100/104/110/112/114/115/117 and discrete outputs 116/118, it is
anticipated that in some systems, the sensors and controls
100/104/110/112/114/115/117/116/118 are signaled through a bus,
wired and/or wireless, such as Inter-Integrated Circuit (I.sup.2C),
car-area network bus (CAN), vehicle area network bus (VAN),
Ethernet, Bluetooth, IEEE 488, USB, FireWire (1394), RS232 bus,
Wi-Fi, etc.
[0046] In general, the persistent storage 88 is used to store
programs, executable code and data such as user financial data in a
persistent manner. The removable storage 90 is used to load/store
programs, executable code, images and data onto the persistent
storage 88. These peripherals are just examples of input/output
devices 80/84/92, persistent storage 88 and removable storage 90.
Other examples of persistent storage include core memory, FRAM,
flash memory, etc. Other examples of removable media storage
include CDRW, DVD, DVD writeable, Blu-ray, compact flash, other
removable flash media, floppy disk, ZIP.RTM., etc. In some
embodiments, other devices are connected to the system through the
system bus 82 or with other input-output connections/arrangements
as known in the industry. Examples of these devices include
printers; graphics tablets; joysticks; and communications adapters
such as modems and Ethernet adapters. Any configuration of
input/output devices is anticipated and the high-efficiency hot
water heater system is not limited to any particular architecture
and/or configuration.
[0047] In some high-efficiency hot water heater systems that
communicate with a central server 40 and/or external information
providers 50, a network interface 80 connects the processor 70 to
the network 10 through a link 78 which is any known network media
such as a cable broadband connection, a Digital Subscriber Loop
(DSL) broadband connection, a T1 line, a T3 line, or a wireless
link such as Wi-Fi, or a cellular data connection.
[0048] Referring to FIG. 4A, a schematic view of a typical
storage-tank hot water heater of the prior art is shown. In this
example of the prior art, cold water enters the tank 1 and is
heated by a heating element 2 that is controlled by a thermostat to
attempt to maintain the temperature of the water in the tank 1 to a
certain temperature (e.g. 125.degree. F.). This example of the
prior art always tries to maintain that certain temperature, even
during the evening or when the covered building is vacant and there
is no demand for hot water. Because it is constantly maintaining
that certain temperature, as heat is lost through the insulation
that typically surrounds the tank 1 and through the supply and
output connections, energy is consumed by the heating element 2,
even though there is no actual use of hot water.
[0049] Referring to FIG. 4B, a schematic view of a typical
on-demand hot water heater of the prior art is shown. In this
example of the prior art, cold water is heated instantaneously by
one or more heating elements 3 that is/are controlled by a flow
sensor such that, as soon as demand for hot water occurs (e.g. a
tap is opened), the heating elements 3 are energized to heat water
from the cold water supply to a desired output temperature.
On-demand hot water heaters of the prior art are typically more
efficient that storage-tank systems as shown in FIG. 4B, but due to
the maximum capabilities of heating elements 3 in raising the
temperature of the cold water supply, when multiple demands are
placed on these on-demand hot water heaters, they often do not
provide hot water that is sufficiently hot. On-demand hot water
heaters will deliver cooler hot water when several people are
filling bathtubs, showering, doing laundry and/or washing dishes.
Additionally, because there is no hot water storage, on-demand hot
water heaters will not supply hot water during a power failure.
[0050] Referring to FIG. 5, a simplified schematic view of the
high-efficiency hot water heater is shown. The exemplary
high-efficiency hot water heater has a hot water storage tank 140
with one or more sources of heat such as heating element(s) 122 (or
main heater) and one or more on-demand heaters 120. Any sources of
heat are anticipated for the heating elements 122 associated with
the tank 140 and for the on-demand heaters 120, including, but not
limited to, electric heating elements, gas flames, oil flames, hot
fluids from a boiler, hot fluids from solar panels, etc.
[0051] Valves 130/132/134 control the flow of water such that the
output (HW) is supplied either directly from the hot water tank
(tank input valve 130 is open, tank output valve 132 is open and
tank bypass valve 134 is closed) or directly from the on-demand
heaters 120 (tank input valve 130 is closed, tank output valve 132
is closed and tank bypass valve 134 is open) or a combination of
both through partial operation of the valves 130/132/134.
[0052] Note that, in some embodiments, there are no internal
heating elements 122 and all water heating is performed, for
example, by the one or more on-demand heaters 120 or provided by
other heat sources such as boilers and solar collectors.
[0053] In this example, cold water from the supply (CW) is heated
by the on-demand heating elements 120 to feed the tank 140 while
the temperature within the tank 140 is maintained by an internal
heating element 122. The valves 130/132/134 and heating elements
120/122 are controlled by the controller based upon the analysis
algorithms 30. For example, in a dormitory situation, at 1:00 AM
(low demand period), the internal heating element 122 maintains a
lower water temperature in the tank 140 and, should demand for hot
water occur, one or more of the on-demand heating elements 120 is
energized and hot water is routed directly from the on-demand
heating elements 120 (tank input valve 130 is closed, tank output
valve 132 is closed and tank bypass valve 134 is open). At a time
later in the morning, based upon a prediction of the analysis
algorithms 30, the internal heating elements 122 are energized to
bring the water temperature in the storage tank 140 to a higher
temperature to meet a high demand as students start to wake.
[0054] Note that in some embodiments, there are no on-demand
heating elements 120 and, therefore, no need for valves
130/132/134. In such a minimal system, the controller 20
predictively controls operation of the internal heating element(s)
122, but such a system lacks on-demand heating elements 120, during
periods where demand is not expected demand results in lower
temperature hot water.
[0055] Referring to FIG. 6, a second schematic view of the
high-efficiency hot water heater is shown. In this example, the
controlled devices 120/122/130/132/134 and sensors/inputs
100S/104S/110S/112S/114S/115S/117S are connected to the controller
20 through a bus 125 to simplify wiring. Alternatively, one or more
of the controlled devices 120/122/130/132/134 and/or sensors/inputs
100S/104S/110S/112S/114S/115S/117S are directly connected to the
outputs 116/118 and inputs 100/104/110/112/114/115/117 of the
controller 20 (e.g. temperature sensor 100S is directly connected
to input 100). In this example, the controller 20 is also connected
to the network 10 (e.g. Internet) for communication with the server
40 and/or one or more information providers 50.
[0056] This exemplary high-efficiency hot water heater has a hot
water storage tank 140 with internal heating element(s) 122 (or
main heater) and one or more on-demand heaters 120. Valves
130/132/134 control the flow of water such that the output (HW) is
supplied either directly from the hot water tank (tank input valve
130 is open, tank output valve 132 is open and tank bypass valve
134 is closed) or directly from the on-demand heaters 120 (tank
input valve 130 is closed, tank output valve 132 is closed and tank
bypass valve 134 is open) or a combination of both through partial
operation of the valves 130/132/134. Operation of the heating
elements 120/122 and valves 130/132/134 is controlled by the
processor 70 of the controller 20 sending and receiving signals
over the bus 125.
[0057] In this example, cold water from the supply (CW) is heated
by the on-demand heating elements 120 to feed the tank 140 while
the temperature within the tank 140 is maintained or increased by
heating element(s) 122 associated with the storage tank 140 (e.g.
internal heating element(s) 122). The valves 130/132/134 and
heating elements 120/122 are controlled by the controller 20 based
upon the analysis algorithms 30 as previously described.
[0058] Note that in some embodiments, there are no on-demand
heating elements 120 and, therefore, no need for valves
130/132/134. In such a minimal system, the controller 20
predictively controls operation of the heating elements 122 that
provide heat to water within the storage tank 140, but such a
system without on-demand heating elements 120 will not provide
properly heated water during periods where demand is not expected
because the water temperature in the storage tank 140 is typically
allowed to decrease during such periods.
[0059] Referring to FIG. 7, a flow chart of a storage-tank hot
water heater of the prior art is shown. In the prior art, the
typical water heater (e.g. 50 gallon home water heaters) has
simple, thermostatically controlled heating elements 2 within the
storage tank 1 (see FIG. 4A). In such heaters, the heating element
is initially de-energized 200, then the following steps repeated
(though this is typically a hardware implementation rather than a
software implementation). The water temperature within the storage
tank 1 is measured 202 and compared to a low-threshold 204. If the
water temperature within the storage tank 1 is below the
low-threshold 204, the heating element is energized 206. The water
temperature within the storage tank 1 is then compared to a
high-threshold 208 and, if the water temperature within the storage
tank 1 is higher than the high-threshold 208, the heating element
is disabled 210. In this exemplary system of the prior art, the
heating elements continue to maintain the water temperature within
the storage tank 1 at a temperature between the low-threshold and
the high-threshold, even when the occupants of the building being
served are not present (e.g. at work, on vacation) or have a low
probability of using hot water (e.g. asleep).
[0060] Referring to FIG. 8, a flow chart of an on-demand hot water
heater of the prior art is shown. In the prior art, the typical
on-demand water heater has a simple, flow-detection controlled
heating element(s) 3 (see FIG. 4B). In such heaters, the heating
element is initially de-energized 220, then the following steps
repeated (though this is typically a hardware implementation rather
than a software implementation). The flow of water is measured 222
and compared to a threshold 224. If the flow of water is above the
threshold 224, the heating element is energized 226. If the flow of
water is below the threshold 224, the heating element is
de-energized 228. In this exemplary system of the prior art, the
heating elements 3 must have sufficient power output to heat the
water as it flows around the heating elements 3. The amount of heat
needed to raise the temperature from ground water temperature to
the desired hot water temperature is directly proportional to the
flow rate and, therefore, if the flow rate exceeds the design
limits of the heating element(s) 3, all users will experience water
that is not as hot as desired. Furthermore, because there is no
storage tank 1 (as in FIG. 4A), if there is a power failure, this
water heater of the prior art will produce no hot water. This is
especially bad if there is a power failure in the morning before
occupants of the affected building leave for work, assuming that
their alarm clocks wake them during the power failure.
[0061] Referring to FIG. 9, a flow chart of an exemplary
high-efficiency hot water heater is shown. This flow chart is
greatly simplified to provide a basic understanding of the
operation of the high-efficiency hot water heater. This example
uses heating elements (e.g. electric heating elements) as the
sources of heat for simplicity purposes. The heating elements
120/122 are initially de-energized 300 then the following steps
repeated: (a) The flow rate is measured 302, then (b) the flow rate
is compared to a first threshold 304 and if greater than the first
threshold THR1 304 (e.g. there is at least some flow of water
indicating demand for heated water), the flow rate is compared 306
to THR2, a second, higher flow rate (i.e. is there low demand or
high demand?). If the flow rate is higher than the THR2 (high
demand), then the on-demand heating element 120 is energized (H=1)
and all valves are open (V1=1, V2=1, V3=1) 308, utilizing any
already heated water from the storage tank 140 to provide the
greatest amount of hot water as possible. If the flow rate is lower
than the THR2 (low demand), then the on-demand heating elements 120
are energized (H=1), the valve between the on-demand heating
elements 120 is opened (V3=1), and the valves leading to and from
the storage tank 140 are closed (V1=0, V2=0) 310 providing
sufficient water heating for the intermediate demand because the
on-demand heating elements 120 are sized to supply the intermediate
demand.
[0062] Next, the history file is consulted 322. The history file
contains, for example, historical usage patterns sorted by
meaningful calendar periods (e.g. days of the week). For example,
an exemplary history file for a dormitory might show the
following:
TABLE-US-00001 Day Hour Usage Monday Midnight-6:00 AM Low Monday
6:00 AM-6:45 AM Moderate Monday 6:45 AM-8:00 AM High Monday 8:00
AM-5:45 PM Low Monday 5:45 PM-11:59 PM Moderate
[0063] The above example only shows one day of the exemplary
history file for brevity purposes. As hot water is used, the
high-efficiency hot water heater updates the history file, with or
without data smoothing to ignore unusually high or low demands. In
this example, there is little or no need to heat the water in the
storage tank 140 between the hours of midnight and 6:00 AM, being
that there is low demand and the on-demand heating elements 120 are
capable of supplying any predicted demand during that period. In
such cases, the need predicted test 324 results in not needed and
the storage tank heater is set to off (I-H=0) 326. At 6:00, it is
predicted that moderate usage will start and last for 45 minutes.
Again, the on-demand heating elements 120 are capable of supplying
any predicted demand during that period, but the analysis
algorithms 30 recognize that starting at 6:45 AM, the demand will
be high (e.g. greater than THR2), so high that the on-demand
heating elements 120 will not be capable of supplying the predicted
demand from 6:45 AM to 8:00 AM. Therefore, the need predicted test
324 determines a high need (flow>THR2) and the storage tank
heater is energized (I-H=1) 328 to preheat the water in the storage
tank 140 to meet the expected high demand period that starts at
6:45 AM. At around 7:40 AM, the need predicted test 324 determines
that sufficient hot water is already available within the storage
tank 140, enough to supplement the on-demand heating elements 120.
The storage tank heating element(s) 122 is again de-energized 328
and the demand for the next 20 minutes or so (until 8:00 AM) is met
by the hot water already in the storage tank 140 and the on-demand
heating elements 122. Because the remainder of the day, only low or
moderate usage is predicted, there is no need to energize the
storage tank heating elements 122 being that the on-demand heating
elements are capable of supplying all predicted demand for that
period.
[0064] The above example is for illustrative purposes and it is
known that a system would generally be much more complicated and
have more or less data.
[0065] Referring to FIG. 10, a schematic diagram of a
neural-network-based high-efficiency hot water heater is shown. The
data available as inputs (ambient 32, schedule data 33, external
data 36, current data 35 and history 34) are examples of input data
for the neural network 430 and it is anticipated that
high-efficiency hot water systems will have more, less, and/or
different inputs. In this example, one or more inputs
32/33/34/35/36 are analyzed by a neural network 430, controlling
the heating elements 120/122 (sources of heat). Any or all of the
inputs 32/33/34/35/36 are used by the neural network 430 to predict
when different amounts of hot water will be needed. For example,
ambient conditions 3 (e.g. weather) are used by the neural network
430 to predict hot water needs based upon internal weighing factors
and other inputs. For example, if previously, users adjusted the
water temperature of their shower dependent upon outside
temperature and humidity, the neural network will "learn" and add
intelligence that uses the ambient data 32 to predict future hot
water demands. In another example, internal ambient conditions tend
to influence a person's shower temperature. For example, when there
is a low humidity in the building, a person tends to take warmer
showers. The neural network 430 monitors such inputs and makes
inferences based upon previous ambient data 32 to predict future
hot water demands.
[0066] Schedule data 33 is also considered by the neural network
430 to predict when demand will occur. For example, in a dormitory,
by knowing weigh values for various aspects of the schedule of
students along with the other data, the neural network 430 controls
the on-demand heating elements 120 and the heating elements 122
associated with the storage tank 140 based upon the data
32/33/34/35/36. With this data, the neural network 430 will learn
that lights out starts at 10:00 PM and classes start at 8:00 AM
and, therefore, the neural network 430 will learn to predict very
low hot water demand after 10:00 PM when the students are asleep,
high hot water usage prior to 8:00 AM when students are waking and
taking showers, and low hot water usage when students are in class
after 8:00 AM, etc.
[0067] Current data 35 is also (optionally) considered by the
neural network 430 to predict when demand will occur. Current data
includes, for example, current flow rates, current hot water output
temperature, and current system input water temperature, etc. The
neural network 430 also uses the current data 35 to determine how
well predictions match actual usage and if any boosting or setback
is needed based upon the current data 35. For example, if low usage
is predicted for the upcoming timeframe but the current data 35
shows that high usage is occurring, then the neural network 430
reacts to such by energizing heating elements 120/122 and
controlling valves 130/132/134 to provide maximum hot water
output.
[0068] Another optional source of data to the analysis algorithms
30 is external data 36. External data includes any data feed that
has the potential to effect hot water usage. This data includes,
but is not limited to, weather predictions, data from an almanac
(e.g. sunrise and sunset times), local news information, school
information, lunch menus, school events, local events, etc. When
such external data 36 is available, the neural network 430 also
considers such in making predictions.
[0069] Although similar to schedule data 33, historic data 34
differs, in that, demand measurements from the past are used by the
neural network 430 to predict future demand. For example, after
installation of the high-efficiency hot water heater system, the
neural network 430 is taught to predict demand based upon on
ambient data 32, schedule data 33, external data 36 and current
data 35, saving historical data and data related to its own
performance in the history data 34. After a number of days of
controlling the heating elements 120/122 and valves 130/132/134
based upon these data 32/33/35/36, the neural network 430 now has
access to historical data 34. For example, one set of 40 students
may have different hot water usage patterns that differ from
another set of the same number of students, perhaps due to
differences in gender and other backgrounds. Initially, the neural
network 430 is taught to make an assumption of the hot water needs
per person, for example, making an average consumption prediction
or a worst-case consumption prediction. As time goes on and history
data 34 is recorded, usage patterns start repeating, and the neural
network 430 consults the historic data 34 to learn how the settings
of the heating elements 120/122 and valves 130/132/134 need be
adjusted based upon the historic data 34. For example, the first
Monday after installation, the neural network 430 has no historic
data 34 and controls the heating elements 120/122 and valves
130/132/134 to preheat the water in storage to a certain
temperature based upon 40 unknown people. By measuring the flow
rates, input water temperature and output water temperature, the
neural network 430 determines that the water in the storage tank
140 does not need to be heated as much as it was, so the next
Monday, the water in the storage tank 140 is heated 10% less and
the flow rates, input water temperature and output water
temperature are again measured to determine how well the
high-efficiency hot water heating system is meeting the demands of
users.
[0070] The neural network 430 accepts data (as described above) and
determines actions (as described above), with the added benefit
that the neural network learns. The neural networks find patterns
in the data as well as filter the data. For a home installation
example, the neural network 430 is initially taught that morning
hot water demand starts at 7:00 AM and ends at 8:00 AM. Over time,
the neural network 430 finds that every Friday morning between 6:00
and 6:15, there is a high demand for hot water and, therefore, will
make adjustments (e.g. learns) to better satisfy that early demand
for hot water. The neural network 430 also filters its data to
ignore irregular data. For example, when users in the prior example
are on vacation and there is no demand for hot water on that day,
the neural network 430 filters out the data from the vacation
period.
[0071] In some high-efficiency hot water heater systems, the basic
neural network software 430 is provided as a package from a
provider of such and is programmed based upon the range of inputs
available to the high-efficiency hot water heater (e.g., schedule,
ambient/weather, history, current data, external data, etc.) to
control the available heating elements 120/122 (see FIGS. 5 and 6)
and/or valves 130/132/134. For some high-efficiency hot water
heater, the neural network software 430 is programmed from
scratch.
[0072] In some high-efficiency hot water heaters, instead of using
true neural networks 430, heuristic algorithms or static logic is
used in the prediction algorithms 30. A simple example in a
dormitory in which all students leave for breakfast and class at
the same time and there are n students registered for that
dormitory, an exemplary heuristic algorithm is: if n is less than
30, preheat the water to t1 at time T1; if n is greater than 30 and
less than 60, preheat the water to t2 at time T2; and if n is
greater than 60, preheat the water to t3 at time T3. In this, the
schedule data 33 is used to determine when the students will be
using the hot water (e.g. before class/breakfast) and how many
students are present. The more students present, the hotter the
water in the storage tank needs to be, therefore, heating starts
earlier and ends when the water reaches this higher
temperature.
[0073] Equivalent elements can be substituted for the ones set
forth above such that they perform in substantially the same manner
in substantially the same way for achieving substantially the same
result.
[0074] It is believed that the system and method as described and
many of its attendant advantages will be understood by the
foregoing description. It is also believed that it will be apparent
that various changes may be made in the form, construction and
arrangement of the components thereof without departing from the
scope and spirit of the invention or without sacrificing all of its
material advantages. The form herein before described being merely
exemplary and explanatory embodiment thereof. It is the intention
of the following claims to encompass and include such changes.
* * * * *