Edited for clarity

When dealing with on-site renewable generation or thermal plants, it is helpful to be able to distinguish which meters use the "source" convention (producing) and which use the "load" convention (consuming). I propose to extend the Haystack standard with the following tags for use with meter equipment:

source // Generates energy
load   // Consumes energy

Note that load is already defined as a chiller point tag, but could be re-used in this context as well.

Specifically, this proposal would apply source and load tags to meter equipment in order to define the sign convention for the measurements:

• Source convention (source tag) defines positive power/energy as power/energy produced (generated)
• Load convention (load tag) defines positive power/energy as power/energy consumed

The load convention is implicit in all present Haystack meters and would be the default in the absence of a tag.

Places where such tags are helpful:

1. Defining functions for virtual submeters or site meters by aggregating physical meters, some of which are generating energy (e.g. from a PV array).
2. Defining rules that check the balance between all thermal sources and loads to ensure minimal leakage or loss, e.g. check that the sum of loads in a chilled water system adds up to approximately the total of all source meters.
3. Defining history functions to compute the proper polarity power or energy for thermal meters based on flow measurements and temperature delta. (E.G. for a chilled water load, entering temp < leaving temp, but for a chilled water source entering temp > leaving temp.)

Is source a generator? Or maybe something a little more generic? Like to understand use cases more

In generic terms, a source is anything that supplies energy and a load is anything that consumes energy. So, PV arrays would be electrical energy sources, while boilers and chillers would be thermal energy sources.

However, I'm proposing use of the tags in a targeted way: to define the sign convention on the associated meter. Source convention (source tag) would define positive power/energy as power/energy produced (generated). Load convention (load tag) would define positive power/energy as power/energy consumed.

I realize now this isn't quite clear in the original post, so I will edit it for clarity.

Use cases...

This becomes relevant if one wants to totalize submeters to check for accuracy against a main meter. Let's say we have elec meters:

Main  = 32 kW
HVAC  = 29 kW
Other = 18 kW
PV    = 15 kW

Where PV is reporting PV energy generated and the others are reporting loads (whole building, HVAC equipment, other circuits). Presumably, we've marked PV, HVAC, and Other as submeters of the main. However, if we want to do a sanity check, we need to subtract the PV, not add it:

29 + 18 + 15 > 32  => Fails check  \\ Wrong: PV has source convention
29 + 18 - 15 = 32  => Passes check \\ Right: PV must be subtracted

Using source and load tags on the meters in this case allows you to write rules (or calculated virtual meters) that properly add the load from all loads and subtract the generation from all sources.

A similar use case is if there are thermal energy meters on (e.g.) all chillers and on all chilled water loads. A good check would be to see if the lost energy exceeds some percentage of the generated energy. The way to do this is to add up the energy from the meters on all chillers (using the source tag to identify them) and subtract all the energy used at the loads (using the load tag to identify them) to get the losses, then check the losses against total chilled water generated.

I supposed another way to do this would be with a convention (or conv?) tag, e.g.

convention:"source"
convention:"load"

I really like this idea. Like you stated it would be very beneficial in several cases. I know it is a little lengthy, but what about:

direction: "production" or "consumption"

In the August, 2016 ASHRAE Journal, there is an interesting article titled "Facility Information Model Standard".

It appears that they are modeling items in the building as

• Load Class (chillers, pumps, air handlers etc.)
• Generator Class (gas powered generators, solar, etc.)
• Meter Class (electric meter and emissions meter)
• Energy Manager Class (things that make energy-related decisions)

If we adopted the tags load and generator for meters, it appears we would align with the ANSI/ASHRAE/NEMA Standard 201-2016 referenced in the article.

Thanks for sharing. Without having read the article, it appears that these are equipment-level designations for the meters? If so, that's different than applying source/load to tags. We're gearing up to work on equip-level tags for electrical equipment and I can take a look at this article to align it with ASHRAE's approach.

FYI: In the newly proposed Haystack tagging convention for points on electrical meters, we decided against source/load style tags altogether in favor of specifying that Haystack will use the load convention for net power and net energy. See this thread for details:

http://project-haystack.org/forum/topic/446

What I have done currently is to create an elecMeterSource tag for an electric meter inside the school building measuring energy output from a generator such as a CHP (combined heat and power) unit. This allows automatic creation of a virtual siteMeter that is a combination of electric meters with tags (1) siteMeter and (2) elecMeterSource.

Automatic processing of the virtual siteMeter results in effortless presentation of the building (1) energy profile, (2) EUI, (3) etc. for buildings that don’t export energy to the grid.

My understanding of Topic 446 is that import / export tags can be applied to meters to address energy exchange with the local utility grid.

The elecMeterSource tag seems in line with the existing elecMeterLoad tag.

From my experience with this school system; I still need a source tag of some kind. Does the new recommendation eliminate the elecMeterLoad tag? I never used it because I always assumed an electric meter inside a building was a load until I came across one that was not.

Sorry; I got busy and never got back to this. Then forgot.

As I understand it, the elecMeterLoad tag (which is a reference, by the way) is to mark individual loads underneath an elec meter. It doesn't formally imply anything about direction of power flow (although implicitly it makes sense that they use the load convention).

In the newly proposed tags (which I hope will be formalized soon), we just state that Haystack uses the load convention for net power and net energy (+ for power consumed, - for power produced). If the meter is backwards from this, then I think to follow the standard you would flip the sign of the data on import.