[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [energyinterop] RE: Error Condition Re: Re: [energyinterop] Observations from the Appliance workshop
I do love the concept of peer-to-peer communications as well. This
said, however, without well defined use cases it would be very difficult to
agree on a design.
So, for instance, is there a use case which allows consumers to schedule
opt in/opt out conditions in conjunction with other preferences? What are exception
use cases? How do alternate sources of energy use cases are handled?
With kind regards,
Further to Ed’s comment, “Using
prices it is clear in most cases that there is no customer economic benefit to
coordinating the use of appliances at least in the US residential markets where
there is no demand charge”; this is absolutely true. The appliance
vendors want to be able to sell into Europe, which does have such a
restriction, and they are likely anticipating (if they’ve though this far about
it) that such a restriction may eventually be seen in some US markets.
I fully agree with Ed’s
elaboration on the vectors-of-prices concept and feel that, combined with
potential demand changes in the future, Jim is right that peer-to-peer
communication of capabilities and present/near-future energy use of appliances
with other appliances is the right thing to address. It’s a concept near
and dear to my heart, being so integral to the LonWorks platform, but it just
makes good sense for things to be aware of their surroundings. Barring
that peer relationship, the ESI/EMS has a great deal to negotiate and while it
could and should handle scenarios, it should be peer-to-peer that handles the
instantaneous (and near-term) occurrence of energy use within devices.
email@example.com [mailto:firstname.lastname@example.org] On Behalf Of James Tillett
Some thoughts I have here is as follows:
An EMS of some type would likely be able to orchestrate a much
more tuned and efficient use of all resources in a facility (whether
residential or commercial), IF it is setup properly. Each environment is
a little different than its neighbor but having a versatile EMS which can
understand and manage its surroundings with appropriately intelligent energy
devices (appliances, HVAC, etc.) should be very effective and getting the most
out of the facility, whatever “the most” happens to be at that location and/or
The downside of this EMS is that for it to be most effective it
probably needs some specific tuning. This may not be such a big deal for
commercial locations because they are used to the commissioning process
anyway. That’s not to say that commissioning is easy, but having a
central location (EMS) from which to manage all resources (which have
appropriate standards to enable cooperation) should make it easier and more
Particularly in a residential environment the typical homeowner
doesn’t want to learn about “commissioning” their house and doesn’t have that
much ambition or desire to go down that road. It would be possible to
build a pretty smart “out of the box” residential EMS, but I’ts still using a
general purpose solution on a specific and unique location.
Here is where intelligent collaboration of devices may be more
valuable. If all appliances are connected to each other, have an
appropriate awareness of each other and their capabilities, an awareness of key
external conditions (weather, DR, etc.), and a rough idea of homeowner
preferences then they could probably take some basic actions between each other
that benefit the home. These could be things like stagger power usage,
precool in preparation of a peak period, and many others. In this way the
appropriate intelligent collaboration would reduce the impact on the homeowner
to be manageable and still provide overall benefit to all stakeholders.
Of course this would be predicated on proper communication
standards, capability, and overall modeling, but in time I think it would work
out. This passive interaction might also be useful to commercial
locations as well.
Hope that makes sense.
James Tillett P.E.
Endeavor Engineering Inc.
From: Ed Cazalet
Let's go back to the last point in your initial email:
"Appliance manufacturers want to hear what it is worth to the grid. I
can give you 5kW for 15 min, 2kW for an hour, 1 kW for 8 hours. What’s it worth
to you? The appliance manufacturers haven’t gotten this data yet. "
Appliance manufacturers may not yet think of it this way, but the simplest
and only way to convey this information is in terms of the current and
forward prices of electric energy. This tells the appliance exactly what it
will cost to run the appliance for any scenario.
And as Sec. Steven Chu said, "the simplest way to communicate consumer
preferences is a green button and a red button on the thermostat, dryer,
etc." The red button says I am willing to pay a lot for comfort /
convenience and the green button says I will give up comfort to save
Given the current and forward prices of power a very simple device can
figure out how much is saved by extending the drying cycle to get a lower
price later or letting the air temperature rise above or below the air
conditioner set point based on the differences between current and forward
prices. Based on which button is set, (red or green ) the device can
decide how to operate.
The actual set point on the thermostat can based on manual setting,
occupancy, iphone control, etc. independent of price. The prices and the
consumer preference button will determine the variability in room temperature
about the set point.
Using prices it is clear in most cases that there is no customer economic
benefit to coordinating the use of appliances at least in the US residential
markets where there is no demand charge.
Note that the prices can reflect both transmission and distribution system
limits and overall energy costs on the grid..
And the prices can change as system conditions change and such changes
broadcast or transmitted to all of the devices.
Prices will be communicated to devices as a vector of prices. Some
devices will only use the current price, other devices will use the current
price and the price 5 min from now or an hour from now and other more intelligent
devices such as air conditioners with ice storage or electric vehicles may look
a vector of prices over several hours
Such price vectors can be determined by tariffs, supply or demand or can be
reliability prices ( critical peak prices) imposed over short periods during
emergencies or peak demand response.
Edward G. Cazalet, Ph.D.
101 First Street, Suite 552
Los Altos, CA 94022
I think what’s missing from your argument is the pivotal role of
customer comfort and his/her preferences. If we remove the customer from the
equations/use cases, then you are 100% correct: some preliminary configurations
on the device and that’s it.
Now, if you add alternative sources of energy in addition to
customer comfort/preferences, then you will either have to have smart devices
that can communicate with one another and with other sources of energy and then
decide what to do OR you could have a higher level of intelligence (ESI) that
implements/orchestrates these scenarios.
In short, I posit that it all depends on how much importance is
given to customers’ comfort and preferences.
With kind regards,
I am thinking of EISA and their interest in collaboration among
appliances, versus EMS central control, versus independent appliance operation.
The Whirlpool rep at the workshop mentioned the need for the washer to talk
with the dryer, and expressed some concern for collaboration between appliances
to limit demand peaks inside the house (although I don’t see why that matters
in the US residential market now, yes for Europe, yes for commercial). From the
grid perspective—who cares what the peak is in your house since it all gets
averaged out over a thousand homes? So, what is the business imperative that
would support all the effort needed to get some distributed mesh appliance
collaboration standard? For EISA, I guess it’s the commercial market.
Thanks Larry for the observations. Let me comment on that.
cake must finish cooking. The oven will not shut off if the price goes up,
unless I tell it it’s worth it to me in general to turn off the oven in the
case of very high prices. But a smart oven will know that a peak is coming and
recommend waiting and give you the option to save money or turn on the oven
must get the clothes dry. So, the dryer defaults to smart saver mode and would
normally delay load start or ask if it can delay. I say NO, or hit the start
button TWICE to tell it I mean NOW.
take a cold shower. So, take one. The smart water heater could learn your
normal usage patterns. We even discussed pre-heating of hot water tanks with
safety features to prevent scald. Maybe even Daikin in doing this already.
house should be cool when I get home. So, I program the thermostat. And I have
some app on my iPhone that can allow updating it from work if I’m early/late,
if I’m nerdy enough.
to me that pool pumps are like the defrost on the fridge—run it off peak
In each case, the interface is at the appliance, or default
built into the appliance. EISA sees a reason for collaboration among
appliances, but I don’t see it really. I see the need for appliances to get
power use off peak, either by shed or shift.
But what does the EMS bring to the table? A single interface to
collect general home preferences maybe. Also, the potential for requiring less
smarts in the individual appliances. But my contention is that removing smarts
from the appliance depends on developing energy profiles for different
appliance classes that somehow communicates all the nuances of how that
appliance could shed or shift or store energy by different amounts in different
combinations for different future price curves. Am I missing something? I guess
EISA members are focused more on the commercial market and have given
significant thought to the sweet spot for energy profiles (amount of detail). I
look forward to getting more details on that. This gets to John’s input below
(sorry—John Who (?) since it bounced, and apparently on energyinterop list as
If I may I would like to add a few comments to this thread.
If we make the assumption that all of these "smart, communicating
devices" are thrown into the application (our homes) without any
preparation for response models then I agree it would be daunting to achieve
the use case that started this thread. But here is where standards could
If we develop a model for typical "smart grid event
types' (that list would not be that long) and a model for logical actions that
devices should take in response to events they may be informed of, (which would
be defined by the equipment mfg based on their knowledge of their equipment and
an understanding of the events) we can create the kind of orchestrated response
All that would need to be communicated to the devices would
be event types or "scenarios" that group events together. Devices
would not have to be in direct peer to peer communication with each other
(although they could be), and the user would not have to be involved in complex
configuration or programming. As an example, when the dryer was installed a
simple menu would be used to select the responses (a short list) that it should
make in response to a defined list of smart grid events. This could include
pricing, or more likely pricing would be abstracted into levels of response
communicated by the defined events.
With this type of a model defined, the communications could
be very compressed binary format capable of being consumed by simple devices
with limited CPU power and using limited bandwidth on low energy networks.
Equipment mfg could implement this model in their devices without ever having
to be aware of the details of any specific end installation.
On Fri, Oct 30, 2009 at 2:05 AM, Michel Kohanim <email@example.com>
I am in total agreement with
you. Further to your point, how would you configure all these isolated devices
to do what you want? Would you go to each individual device and set them up
individually? What if you are not home? Does this then mean that each
device/appliance should have a user interface? Should all be schedule aware?
Should they be able to communicate amongst one another? If so, what is the
minimum set of communications constructs?
When EISA talks about Zero Net,
one has to also consider alternate sources of energy and thus it would be quite
a daunting task to have independent devices act in unison (and based on
customer preferences) to achieve such goals.
With kind regards,
Michel Kohanim, C.E.O
Universal Devices, Inc.
Thanks for the
observations. For fun, let’s suppose you have several of the items mentioned
below: AC, pool pump, dryer, hot water, and oven, and some friends are coming
over for dinner. You might say:
Seems to me hard
for any automated system to do, let alone an isolated appliance with limited
But if I had a
home EMS, I could tell the EMS provided the appliances used U-SNAP or another
method of communication to the EMS to receive instructions rather acting in
From: Holmberg, David [mailto:firstname.lastname@example.org]
I attended the EPRI “DR-Ready Appliance Workshop” yesterday
in Knoxville, and the discussion ranged around appliance communications and
action for demand response.
Basic agreement on what “smart grid ready” means:
Can understand SEP (or, as I observed, some standard data syntax and
semantics and transport)—EPRI and U-SNAP are pushing for a standard connector
that would allow plugging in an external comms chip. But you still need the app
Of course, how you accomplish the above for specific device classes (in
different regions, see below) might need some definition when it comes time to
do compliance testing. And what kind of signal are you going to feed an
appliance to prove it can shed/shift, and how much? Maybe you need a standard
forward price curve representative of different kinds of typical peaks. Maybe
you need a standard forward mode signal, similar to what the DRAS feeds to the
Simple Client in OpenADR.
Another model that was advocated (not by the appliance manufacturers nor
by me) is having more/most of the intelligence at some EMS and passing a
simpler signal to the appliance. To me, this requires communications from the
appliance to the EMS (at least a standard energy profile) plus it requires a
I realized that there are perhaps limits on how universal appliances can
be. DR programs have very real differences in different utility territories due
to very real weather and regulatory differences. AC is all that matters in
Phoenix (besides turning off pool pumps)—hot water is not an issue because
water comes into the house hot. That won’t be the case in some other places.
This came up relative to the question of whether appliance loads really matter.
But a dryer and oven draw more power than the AC. And a refrigerator only draws
70W on average, but I have 3 and they run 24/7. So, how much does the
regionality issue affect energy management algorithms for appliances? A rep
from Daikin says they ship products that can go anywhere, just need the right
software update to get tuned control algorithms.
Appliance manufacturers want to hear what it is worth to the grid. I can
give you 5kW for 15 min, 2kW for an hour, 1 kW for 8 hours. What’s it worth to
you? The appliance manufacturers haven’t gotten this data yet.
NIST Building and Fire Research Lab
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]