Re: [802.3_4PPOE] Power price index update

Thread Links	Date Links
Thread Prev	Thread Next	Thread Index	Date Prev	Date Next	Date Index

Bruce & all –

This email has two parts – one about Bruce & Geoff’s discussion and one about what Lennart actually proposed.

Part 1: The proposal, actionable for 802.3bt:

What Lennart actually proposed is, IMO, an encoding, not a description of how or why to use the field. I think the encoding is useful and I think a nonlinear mapping (rather than the current linear mapping) is useful. However, the text preceding it is problematic, because it refers to the term “price” in numerous places, something we scrupulously avoided in the initial text.

79.3.8.2 PSE power price index

The PSE power price index field shall contain an index of the current price of electricity compared to what the PSE

considers the nominal electricity price. The determination of the nominal electricity price is implementation dependent.

The field is encoded as defined in Equation (79–1). The PSE sets the value of this field taking the availability of power

from any external and internal resources, and the relative supply and demand balance, into account. A value of 0xFFFF

means that no power price index is available

The original text, while calling the field “power price index” only uses the term “price” in connection with “index”.

It describes the field as representing, “a linear index of the current value of electricity” which keeps us clean of the line for “cost”, “price”, etc.

I recommend staying closer to the original wording, and modifying Lennart’s first paragraph and the text after where KPPI as below, to relate to value rather than price.

The PSE power price index field shall contain an index of the current ~~price~~ value of electricity compared to what the PSE

considers the nominal value~~electricity price~~. The determination of the nominal electricity value~~price~~ is implementation dependent.

The field is encoded as defined in Equation (79–1). The PSE sets the value of this field taking the availability of power

from any external and internal resources, and the relative supply and demand balance, into account. A value of 0xFFFF

means that no power price index is available

… (table and equation as in Lennart’s)…

where

KPPI is the power price index expressed as a factor ranging from 0.0004 to 100 the nominal value~~price~~

-------

Part 2:

Now to Bruce & Geoff’s discussion:

Bruce, you know that I support the concept of this index – I believe it leads us to an infrastructure capable of managing energy. We need, however, to figure out the right place for it. 802.3bt is the right place to define the field. It isn’t however, the right place for tutorial text on how to use it, or informative text about what it means.

If the information about how to use it is necessary for interoperable power-price systems, then I believe that functionality is a higher-layer function (at least a layer 2 exchange, with some layer 3-4 attributes), and needs to be specified in another (or its own) standard, most likely above the 802 level. This (other) standard could give the hard requirements necessary. That standard can then be (later) referenced by 802.3

If the information isn’t necessary for interoperable power-price systems, then, it is only one possible use. Application whitepapers can be written by other bodies (including IEEE 802 coordinating groups like those in 802.24 for smart grid and IoT) describing what these mean and how they are used. You could also consider defining a reference model within the Ethernet Alliance or a similar group (my preference would be the EA, because most of the PoE-related folks are involved and then would be aware).

-george

From: Bruce Nordman [mailto:bnordman@xxxxxxx]
Sent: Wednesday, October 25, 2017 11:33 AM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] Power price index update

Geoff--

I would be happy to fill you in on the background for this by phone.

I think would be a lot more time-efficient than email.

The key for this field is that it is to communicate the relative supply/demand

balance for electricity on the link, as measured by a local price. This is a

different concept (and have different values) than something like a utility grid

AC power price at the meter (there are application-layer protocols for

communicating these prices over many hops). So, this is to be only about

the link.

There should be informative text somewhere to provide guidance on

how it is recommended that the PSE and PD use this field. I was planning

on writing such text, drawing on considerable material I have on this topic

available on my web page, once I figure out the location, format, suitable

length, etc.

Thanks,

--Bruce

On Tue, Oct 24, 2017 at 11:36 PM, Geoff Thompson <thompson@xxxxxxxx> wrote:

Lennart-

It is our job to provide the field that provides the raw stats from PoE.

Doing any higher level calculation or comparison is out of scope.

Geoff

On Oct 24, 2017, at 10:58 PMPDT, Lennart Yseboodt <lennartyseboodt@xxxxxxxxx> wrote:

Hi Geoff,

This field has been included in the draft for almost 2 years. I am improving the definition, not adding a new field.

This was originally proposed by Bruce Nordman of LBNL such that PDs can participate in load shedding schemes.

Much more definition is indeed needed to determine how PSEs would set the index value and how PDs should react to this. That for sure is out of scope for 802.3, but we have to provide a field in order to be able to convey the data.

Kind regards,

Lennart

On 25 Oct 2017 7:53 AM, "Geoff Thompson" <thompson@xxxxxxxx> wrote:

Lennart-

I would assert that this is out of scope for P802.3bt.
It belongs in a higher level management project.

Best regards,

Geoff

> On Oct 24, 2017, at 7:10 AMPDT, Lennart Yseboodt <lennartyseboodt@xxxxxxxxx> wrote:
>
> Hello,
>
> Please find attached an update to the PSE Power price index
> specification. The attached text better specifies how to interpret the
> field and has an optimized mapping between the LLDP field and the
> actual index.
>
> Kind regards,
>
> Lennart<yseboodt_01_1117_powerpriceindex.pdf>

Bruce Nordman
Lawrence Berkeley National Laboratory
nordman.lbl.gov
BNordman@xxxxxxx
510-486-7089
m: 510-501-7943