For an electric meter, the demand (power) is clearly tagged with kw. However, I don't see any analagous tag (e.g. power or demand or energyRate) for a thermal meter (e.g. a waterMeter measuring thermal demand within a chilled water system).
For example, if I have a flow meter and a temperature sensor for the output of a chiller, I would clearly use something like
chilled water leaving temp sensor
chilled water entering temp sensor
chilled water flowRate sensor
...and possibly also include a delta. However, if I want to tag a thermal demand (power) point (either measured or computed) in, e.g., BTU/h, what would I use?
What are others using? Is there a proposal for a standard tag?
Jason BriggsThu 22 May 2014
Stephen... We don't have a proposal for this yet. Let's make this thread a proposal. Can you formalize exactly what you would like to see. Maybe take some more examples. I think this is important that we get this nailed down soon.
Stephen FrankThu 22 May 2014
Well, I'm very new to Haystack compared to most of you on this forum, but the proposal would be to have one or several tags which designate a thermal energy measurement for a water or steam meter. (Although, steamMeter doesn't seem to be a formal tag.)
The common use case is a BTU meter which tracks hot or chilled water consumption for a piece of equipment, a subsystem, or an entire building (in the case of campuses with centralized hot/chilled water distribution systems). The meter might also track production.
I propose either power (analogous to flow) or energyRate (analogous to flowRate) to mark points which measure/report thermal power. Such points are related to the product of the temperature delta and the flow rate by the specific heat and density of the fluid. Similarly, energyConsumption (analogous to flowConsumption) could be used for a totalized thermal energy value, if desired.
By the way, what is the difference between flow and flowRate in this context?
Brian FrankThu 22 May 2014
Although, steamMeter doesn't seem to be a formal tag.)
The fooMeter pattern defined here implicitly defines the steamMeter tag. I haven't mapped all those into the tag reference, but we can do that if everyone thinks it will help?
By the way, what is the difference between flow and flowRate in this context?
For metering the intention was to use flowRate and flowConsumption to correspond to kW demand and kWh consumption respectively.
For this case, the terms "power" and "energyConsumption" sound a little too generic in my opinion. How about more explicit terms such as thermalRate and thermalConsumption?
Stephen, can you provide examples of what kind of units might be used for these two points in a steam meter, chilled water meter, etc? Some concrete use cases would really help me understand since I am not familiar with these terms
Stephen FrankFri 23 May 2014
The specific use case I'm thinking of is a BTU meter on the output of a chiller or boiler plant, typically calculated from a measured volumetric flow rate and a measured temperature delta. Here's an arbitrary example of such a meter that provides a totalized output. On our campus, we also have BTU meters on the input lines to specific buildings to track thermal energy consumption from our centralized hot and chilled water systems.
Typical units for measured power would be BTU/h, kBTU/h, or possibly tonrefh (tons of refrigeration per hour). I assume SI systems would report in kW.
Often, one can get inlet and outlet temperature (°F) and volumetric flow rate (gal/min) out of the same meter.
Jacob FrySun 25 May 2014
Hi All,
We have encountered this situation with thermal meters at a shopping centre. The building owner was charging tenants for chilled water using delivered kWh (thermal).
One way to approach this could be to tag the meters with thermalMeter, as analogous to elecMeter. I think the units could still be kWh, since the fact that we are measuring heat flow and not electricity is inherent in the thermalMeter tag.
This issue also arises on chillers where a load point could refer to either kW (electric) or kW (thermal). So a thermal tag might be appropriate here.
In this context, sometimes thermal is differentiated from refrigerative (i.e. kWth versus kWr), however in my view thermal is the more generic term.
Hope this is constructive.
Richard McElhinneyMon 26 May 2014
Jacob,
in our work on Cooling and Heating plants it has become apparent that it is important to distinguish between chilled water as thermal energy and hot water as thermal energy. So a thermal tag may be appropriate but I think that should be qualified with another tag indicating the type of thermal energy.
We already have chilled and heating so these could be reused in this instance.
Furthermore, for the units we came up with the idea of kilowatt refrigeration hours ( kWrh ) as a representation for chilled thermal energy consumed. I'm not sure this is an official SI unit anywhere, however in our work we needed to distinguish between electrical consumption ( kWh ) and chilled thermal energy consumption in a plant.
In our application we also need to come up with some unit to further distinguish heating thermal energy consumption...but we haven't really put much thought into that.
I guess overall I'm saying that I would prefer not to be overly generic in this case. I think the modelling is simple and complete enough to tolerate the distinction between the type of thermal energy being produced.
Cheers, Richard
Brian FrankTue 27 May 2014
Good ideas, how do we narrow this down to a real proposal? I am not sure I fully grasp the various options. To restate what I think we need to do:
We already have standardized tags for modeling a steamMeter or chilledWaterMeter (question: or do we care about adding these points to other types of equipment?)
What we need is standard tags for the demand and consumption points as a measure of energy
Standardized unit or units for these two points
If they are truly are energy points that can be respresented as kw and kwh, should we reuse those? Maybe just an additional tag such as thermal?
Stephen FrankWed 28 May 2014
Given your comments about xMeter being a standardized format and other input above, I would propose a thermal tag to identify thermal power/energy measurements, plus one of the following options:
power and energy tags to designate power and energy measurements, respectively, or
energyRate and energyConsumption tags to designate power and energy measurements, respectively (analogous to flowRate and flowConsumption).
kw and kwh are units as well as tags and, to me at least, really only make sense as tags in the context of electricity. However, even there it would be possible to have a kw point with units of MW, for instance, which can be a little confusing.
Under this proposal, a waterMeter equip that measures chilled water demand might have tags:
chilled waterMeter
with points:
chilled water entering temp sensor
chilled water leaving temp sensor
chilled water flowRate sensor
chilled water temp delta sensor
chilled water thermal power sensor
Richard McElhinneyThu 29 May 2014
Definitely getting closer.
I like @Stephen's approach with the energyRate and energyConsumption tags. As per my previous post I also like the idea of tagging a meter (for this use case) as being either chilled or heating.
However, I would propose to leave the waterMeter designation as is, and go with the thermalMeter + chilled or heating. Also I'm not sure about having temperature points associated with a water meter as typically they are physically separate sensors or part of the machine (chiller) available on a high level interface.
So taking the straw-man approach my proposal would be this:
id:@meter1
thermalMeter
chilled
siteMeter (or @submeterOf if there is a thermal metering hierarchy)
equip
siteRef: @mySite
...
id:@meter1-flow-point
sensor
point
energyRate
equipRef: @meter1
unit: "kW"
...
id:@meter1-consumption-point
sensor
point
energyConsumption
equipRef:@meter1
unit:"kWh"
...
In this instance I'm Ok with keeping kWh and kW for the units as the type of energy is implicit in the designation of the parent meter equipment rec. That way we can keep some consistency for the units of energy consumption and instantaneous energy rate.
Brian FrankThu 29 May 2014
Meter Level
What to tag the meter:
Richard's proposal is thermalMeter + ( chilled or heating )
I think I'd prefer chilledWaterMeter, hotWaterMeter, or steamMeter
To me waterMeter is for measuring water consumption. This sort of meter is measuring energy between entering and leaving temperatures correct?
Point Level
Tags to use for demand and consumption:
power and energy tags
energyRate and energyConsumption
I think the terms power and energy are actually the best since they are formally defined by physics. But then again maybe not good for this context.
One aside to this, should whatever we define also be applied to an electric meter's kw and kwh tags? In past there has been some concern about these two special tags using a unit and not being generic such as demand and consumption.
Jason BriggsFri 13 Jun 2014
We really need to nail this down. I suggest with Meters we do this.
Every type of meter should have a meter tag on it. Now that the amount of meter types is growing, you should for sure be able to find all your meters without having to search for each meter individually.
To do make up water, you would add 2 tags waterMeter and makeUpWater
It's important to know that a chilledWaterMeter is not a waterMeter. That is what is a little confusing to me. Cause you are actually measuring water too, but you are using it to gather energy data. So my question is this, if you are literally just measuring gpm on chilledWater should you use the waterMeter tag or the chilledWaterMeter tag?
Brian FrankFri 13 Jun 2014
Good point about meter. Although technically every meter regardless of type should have siteMeter or submeterOf so you could do things that way too.
But I could get on board with the meter tag in which case we'd be going down the path where we did with dischargeTemp into dischage air temp sensor. So would we redefine elecMeter to be elec meter?
Jason BriggsFri 13 Jun 2014
I personally think breaking out elec and meter makes much more sense. It's just a major change that we should do soon if we do do it. I'm also ok with leaving it the same, just cause so many people have already tagged it that way. Meaning, I think if I have the elecMeter and meter that is good enough. It makes things easy and doesn't create a big change.
More importantly I think we need to all agree on
New Meter Tags to Add
chilledWaterMeter
hotWaterMeter
condenserWaterMeter
steamMeter
Let's put it to vote
Stephen FrankMon 16 Jun 2014
I personally prefer the individual tag approach, e.g. chilledwatermeter, but I can get behind any standard system.
The downside of breaking out individual tags is that exclusionary boolean algebra becomes more difficult. E.G., if you want only regular (not chilled or hot) water meters, you have to search for:
water and meter and not (chilled or hot)
vs. just:
waterMeter
This tradeoff is worth it, in my opinion.
If adding specific meter tags instead, then the list provided by Jason seems reasonable.
Nicholas HarkerMon 16 Jun 2014
I am also a fan of the individual tag approach, but I feel that we should avoid the exclusionary boolean algebra that Stephen noted above. To do this, we would need to define an exclusive tag combination for water meter, as we have for the others. After talking with our engineering group, I'd like to propose the following:
Water:
Water (Flow) Meter - water flow meter
Chilled Water Meter - water chilled meter
Condenser Water Meter - water condenser meter
Hot Water Meter - water hot meter
Steam:
Steam Meter - steam flow meter
Steam Condensate Meter - steam condensate meter
Electric:
Electric Meter - elec meter
Natural Gas:
Gas Meter - gas meter
The main additions to this proposal are the flow tag for both steam and non-thermal water meters and a new steam condensate meter definition.
I recognize this would represent a breaking change, but we would be willing to update existing projects to see this become the standard.
Anthony RovanoMon 16 Jun 2014
While there are benefits to both individual and concatenated tag approaches, the utilities themselves are not the same as the sums of their individual parts.
chilledWaterMeter(s) measure volume and rate of recirculation, vs waterMeter, elecMeter, and gasMeter which each measure the introduction or usage of that utility type.
The exclusionary boolean algebra mentioned by Stephen would be mandatory for any water based query, a total water query would always provide undesirable results in any environment with either a chilled or heating hot water system.
water hot meter best references domestic hot water, where heatingHotWater would be for radiant heating or other closed-loop systems.
The approach that takes advantage of individual tag flexibility with concactenated specificity would be to use water, elec, and gas with the tag of meter, and add chilledWater and heatingHotWater as the utility type on equipment and points where appropriate.
Denis OConnorTue 17 Jun 2014
In addition to the consideration for
makeUpWater (steam boiler systems and cooling towers)
that Jason mentioned earlier;
Keep in mind
feedWater (steam boilers)
and
blowDownWater (cooling towers and steam boilers)
I am not voting on the individual or concatenated approach, just wanting to get all of the meters on the table.
Chris OlmstedTue 17 Jun 2014
I'm in the individual tag camp and here's why:
I believe that drilling down to more specific points should be as simple as adding additional tags to a query. There will be exceptions, like blowDown or makeUp, where a concatenated approach is necessary because the individual tags are meaningless (blow and down). Individual tags will make haystack easily expandable and flexible. If every tag combination has a unique concatenated name, like primaryChilledWaterFlowMeter, then we might as well use point names from the control system.
As for the potential for exclusionary boolean algebra: why not add more tags? From the example above:
if you want only regular (not chilled or hot) water meters, you have to search for: water and meter and not (chilled or hot)
I'm not sure what a regular water meter is, but let's say it's a domestic meter, then you can use:
water and meter and domestic
I believe anytime we have a case where a set of points can only be accessed though exclusionary logic, that is a sign that we need to add tags to the library.
Brian FrankWed 18 Jun 2014
Okay, if everyone is willing to make a breaking change then my vote is to go individual tag route. That is consistent with what we've done with points dischageTemp went to discharge air temp sensor.
Going thru previous posts, this what I have for full list of meter types:
water flow meter
water chilled meter
water condenser meter
water hot meter
water makeup meter
water blowdown meter
water feed meter
steam flow meter
steam condensate meter
elec meter
gas meter
How does everyone vote for individual tags versus compound tags?
Any other meter types we should capture while we are it?
Keith BishopWed 18 Jun 2014
I think the individual tag model is where we should go. I do think we should discuss a transition plan for this change.
If we are considering splitting out combined meter tagging, should we also look at the other combined tags and discuss their separation.
As far as the meter list above, I don't know that "flow" is a tag that we want to adopt on these fluid meters. Fluid metering is measuring flow. With the water meter, should we use the tag "domestic" instead? The steam "flow" tag should be able to be dropped. Again, a meter on a steam system is going to measure flow. There shouldn't be a need to differentiate it from other sensor type objects.
The "steam" tag should be able to be dropped from the condensate meter. The "condensate" tag should indicate a piece of a system that represents a portion where a phase change has happened from the steam portion of the system. I understand that in many cases condensate is measure instead of steam flow to more accurately measure the mass flow rate of the system to avoid the inherent density compensation issues with measuring steam. No matter the underlying intent, it is still a condensate meter.
I don't know that we have a good use case at this time but as this model grows I could see the need for other metering types for things such as pressurized air systems or specialized water systems such as deionized water. With that said, I think the current list represents the majority of employed meters.
One last thing. With this discussion about metering I think we should constrain the use of "meter" to objects that either measure a rate of use or a totalized consumption.
John SublettWed 18 Jun 2014
I like the individual tag approach. It makes for much flexible queries and I think it should be used consistently throughout the model.
If you are considering a breaking change like breaking "elec" apart from "meter", how about fixing the energyRate and energyConsumption tags to be consistent too and removing the kw and kwh tags? Units should be specified in the "unit" tag.
In fact, in the interest of individual tags, shouldn't "flow", "energy", "rate", and "consumption" all be individual tags?
Keith BishopWed 18 Jun 2014
Can we assume that metered energy points are "rate" points unless a hisTotalized tag is used? Or should we use the unit tag to make this differentiation (kW or gpm are rates whereas kWh and gal are consumption)?
Richard McElhinneyWed 18 Jun 2014
@Brian, I'm on board with the individual tag approach as well. It is definitely more consistent with the other models we've got. So +1 for me.
In terms of other meters, now that we've (basically) agreed on the individual tag approach it should be easier to capture new meters so perhaps it's worthwhile moving forward with what you have proposed.
WRT what @John Sublett says, he raises a good point in proposing to break up other compound tags. While we are making a breaking change here is it better to "rip the bandaid off quickly" and review the rest of the current tags to see if there is areas where we can encourage consistency.
This might be too much of a breaking change to bite off in one go, but just throwing it out there for discussion.
For example a quick look at the documentation reveals some potential for further consistency:
I guess this kind of hijacks the thread and maybe this should be broken out into another topic so it doesn't get confused with metering conversation.
However, I would be interested in feedback.
Zach SchweinfurthFri 20 Jun 2014
After reading through this thread, I want to express my support for the individual tag approach. I think Chris Olmsted is correct in saying that if we build up unique compound tags for many things, then we should just use point names from control systems. Moreover, I agree with Chris that in some cases, compound tags might be necessary when an atomic tag means nothing really by itself. I don’t think we should be so rigid with the atomic tags that we don’t add compound tags in when necessary.
A general guiding rule(s) could be something like: Tags should contain no more words than necessary to identify the underlying concept. And if a compound concept can be equally conceptually represented by two (or n) atomic tags, then the atomic tags should be used over the compound tag provided that each atomic tag represents an engineering concept of importance on its own.
So taking the electric meter example, electricity and meters are conceptually separate things that have engineering significance on their own. And when combined, the tags create a concept distinct from their parts, which is also of significance to engineering.
Stephen Frank Wed 21 May 2014
For an electric meter, the demand (power) is clearly tagged with
kw
. However, I don't see any analagous tag (e.g.power
ordemand
orenergyRate
) for a thermal meter (e.g. awaterMeter
measuring thermal demand within a chilled water system).For example, if I have a flow meter and a temperature sensor for the output of a chiller, I would clearly use something like
...and possibly also include a delta. However, if I want to tag a thermal demand (power) point (either measured or computed) in, e.g., BTU/h, what would I use?
What are others using? Is there a proposal for a standard tag?
Jason Briggs Thu 22 May 2014
Stephen... We don't have a proposal for this yet. Let's make this thread a proposal. Can you formalize exactly what you would like to see. Maybe take some more examples. I think this is important that we get this nailed down soon.
Stephen Frank Thu 22 May 2014
Well, I'm very new to Haystack compared to most of you on this forum, but the proposal would be to have one or several tags which designate a thermal energy measurement for a water or steam meter. (Although,
steamMeter
doesn't seem to be a formal tag.)The common use case is a BTU meter which tracks hot or chilled water consumption for a piece of equipment, a subsystem, or an entire building (in the case of campuses with centralized hot/chilled water distribution systems). The meter might also track production.
I propose either
power
(analogous to flow) orenergyRate
(analogous to flowRate) to markpoint
s which measure/report thermal power. Such points are related to the product of the temperature delta and the flow rate by the specific heat and density of the fluid. Similarly,energyConsumption
(analogous to flowConsumption) could be used for a totalized thermal energy value, if desired.By the way, what is the difference between
flow
andflowRate
in this context?Brian Frank Thu 22 May 2014
The fooMeter pattern defined here implicitly defines the
steamMeter
tag. I haven't mapped all those into the tag reference, but we can do that if everyone thinks it will help?For metering the intention was to use flowRate and flowConsumption to correspond to kW demand and kWh consumption respectively.
For this case, the terms "power" and "energyConsumption" sound a little too generic in my opinion. How about more explicit terms such as
thermalRate
andthermalConsumption
?Stephen, can you provide examples of what kind of units might be used for these two points in a steam meter, chilled water meter, etc? Some concrete use cases would really help me understand since I am not familiar with these terms
Stephen Frank Fri 23 May 2014
The specific use case I'm thinking of is a BTU meter on the output of a chiller or boiler plant, typically calculated from a measured volumetric flow rate and a measured temperature delta. Here's an arbitrary example of such a meter that provides a totalized output. On our campus, we also have BTU meters on the input lines to specific buildings to track thermal energy consumption from our centralized hot and chilled water systems.
Typical units for measured power would be BTU/h, kBTU/h, or possibly tonrefh (tons of refrigeration per hour). I assume SI systems would report in kW.
Often, one can get inlet and outlet temperature (°F) and volumetric flow rate (gal/min) out of the same meter.
Jacob Fry Sun 25 May 2014
Hi All,
We have encountered this situation with thermal meters at a shopping centre. The building owner was charging tenants for chilled water using delivered kWh (thermal).
One way to approach this could be to tag the meters with
thermalMeter
, as analogous toelecMeter
. I think the units could still be kWh, since the fact that we are measuring heat flow and not electricity is inherent in thethermalMeter
tag.This issue also arises on chillers where a
load
point could refer to either kW (electric) or kW (thermal). So athermal
tag might be appropriate here.In this context, sometimes thermal is differentiated from refrigerative (i.e. kWth versus kWr), however in my view thermal is the more generic term.
Hope this is constructive.
Richard McElhinney Mon 26 May 2014
Jacob,
in our work on Cooling and Heating plants it has become apparent that it is important to distinguish between chilled water as thermal energy and hot water as thermal energy. So a
thermal
tag may be appropriate but I think that should be qualified with another tag indicating the type of thermal energy.We already have
chilled
andheating
so these could be reused in this instance.Furthermore, for the units we came up with the idea of
kilowatt refrigeration hours
(kWrh
) as a representation for chilled thermal energy consumed. I'm not sure this is an official SI unit anywhere, however in our work we needed to distinguish between electrical consumption (kWh
) and chilled thermal energy consumption in a plant.In our application we also need to come up with some unit to further distinguish heating thermal energy consumption...but we haven't really put much thought into that.
I guess overall I'm saying that I would prefer not to be overly generic in this case. I think the modelling is simple and complete enough to tolerate the distinction between the type of thermal energy being produced.
Cheers, Richard
Brian Frank Tue 27 May 2014
Good ideas, how do we narrow this down to a real proposal? I am not sure I fully grasp the various options. To restate what I think we need to do:
If they are truly are energy points that can be respresented as
kw
andkwh
, should we reuse those? Maybe just an additional tag such asthermal
?Stephen Frank Wed 28 May 2014
Given your comments about
xMeter
being a standardized format and other input above, I would propose athermal
tag to identify thermal power/energy measurements, plus one of the following options:power
andenergy
tags to designate power and energy measurements, respectively, orenergyRate
andenergyConsumption
tags to designate power and energy measurements, respectively (analogous toflowRate
andflowConsumption
).kw
andkwh
are units as well as tags and, to me at least, really only make sense as tags in the context of electricity. However, even there it would be possible to have akw
point with units of MW, for instance, which can be a little confusing.Under this proposal, a
waterMeter
equip that measures chilled water demand might have tags:with points:
Richard McElhinney Thu 29 May 2014
Definitely getting closer.
I like @Stephen's approach with the
energyRate
andenergyConsumption
tags. As per my previous post I also like the idea of tagging a meter (for this use case) as being eitherchilled
orheating
.However, I would propose to leave the
waterMeter
designation as is, and go with thethermalMeter
+chilled
orheating
. Also I'm not sure about having temperature points associated with a water meter as typically they are physically separate sensors or part of the machine (chiller) available on a high level interface.So taking the straw-man approach my proposal would be this:
In this instance I'm Ok with keeping
kWh
andkW
for the units as the type of energy is implicit in the designation of the parent meter equipment rec. That way we can keep some consistency for the units of energy consumption and instantaneous energy rate.Brian Frank Thu 29 May 2014
Meter Level
What to tag the meter:
thermalMeter
+ (chilled
orheating
)chilledWaterMeter
,hotWaterMeter
, orsteamMeter
To me
waterMeter
is for measuring water consumption. This sort of meter is measuring energy between entering and leaving temperatures correct?Point Level
Tags to use for demand and consumption:
I think the terms
power
andenergy
are actually the best since they are formally defined by physics. But then again maybe not good for this context.One aside to this, should whatever we define also be applied to an electric meter's
kw
andkwh
tags? In past there has been some concern about these two special tags using a unit and not being generic such asdemand
andconsumption
.Jason Briggs Fri 13 Jun 2014
We really need to nail this down. I suggest with Meters we do this.
Every type of meter should have a meter tag on it. Now that the amount of meter types is growing, you should for sure be able to find all your meters without having to search for each meter individually.
Meter Level
Existing Meter Tags
New Meter Tags to Add
chilledWaterMeter
hotWaterMeter
condenserWaterMeter
steamMeter
Some Meter Combinations
To do make up water, you would add 2 tags waterMeter and
makeUpWater
It's important to know that a chilledWaterMeter is not a waterMeter. That is what is a little confusing to me. Cause you are actually measuring water too, but you are using it to gather energy data. So my question is this, if you are literally just measuring gpm on chilledWater should you use the waterMeter tag or the
chilledWaterMeter
tag?Brian Frank Fri 13 Jun 2014
Good point about
meter
. Although technically every meter regardless of type should havesiteMeter
orsubmeterOf
so you could do things that way too.But I could get on board with the
meter
tag in which case we'd be going down the path where we did withdischargeTemp
intodischage air temp sensor
. So would we redefineelecMeter
to beelec meter
?Jason Briggs Fri 13 Jun 2014
I personally think breaking out
elec
andmeter
makes much more sense. It's just a major change that we should do soon if we do do it. I'm also ok with leaving it the same, just cause so many people have already tagged it that way. Meaning, I think if I have the elecMeter andmeter
that is good enough. It makes things easy and doesn't create a big change.More importantly I think we need to all agree on
New Meter Tags to Add
chilledWaterMeter
hotWaterMeter
condenserWaterMeter
steamMeter
Let's put it to vote
Stephen Frank Mon 16 Jun 2014
I personally prefer the individual tag approach, e.g.
chilled
water
meter
, but I can get behind any standard system.The downside of breaking out individual tags is that exclusionary boolean algebra becomes more difficult. E.G., if you want only regular (not chilled or hot) water meters, you have to search for:
vs. just:
This tradeoff is worth it, in my opinion.
If adding specific meter tags instead, then the list provided by Jason seems reasonable.
Nicholas Harker Mon 16 Jun 2014
I am also a fan of the individual tag approach, but I feel that we should avoid the exclusionary boolean algebra that Stephen noted above. To do this, we would need to define an exclusive tag combination for water meter, as we have for the others. After talking with our engineering group, I'd like to propose the following:
Water:
Steam:
Electric:
Natural Gas:
The main additions to this proposal are the
flow
tag for both steam and non-thermal water meters and a new steam condensate meter definition.I recognize this would represent a breaking change, but we would be willing to update existing projects to see this become the standard.
Anthony Rovano Mon 16 Jun 2014
While there are benefits to both individual and concatenated tag approaches, the utilities themselves are not the same as the sums of their individual parts.
chilledWaterMeter(s) measure volume and rate of recirculation, vs waterMeter, elecMeter, and gasMeter which each measure the introduction or usage of that utility type.
The exclusionary boolean algebra mentioned by Stephen would be mandatory for any water based query, a total water query would always provide undesirable results in any environment with either a chilled or heating hot water system.
water hot meter best references domestic hot water, where heatingHotWater would be for radiant heating or other closed-loop systems.
The approach that takes advantage of individual tag flexibility with concactenated specificity would be to use water, elec, and gas with the tag of meter, and add chilledWater and heatingHotWater as the utility type on equipment and points where appropriate.
Denis OConnor Tue 17 Jun 2014
In addition to the consideration for
makeUpWater (steam boiler systems and cooling towers)
that Jason mentioned earlier;
Keep in mind
feedWater (steam boilers)
and
blowDownWater (cooling towers and steam boilers)
I am not voting on the individual or concatenated approach, just wanting to get all of the meters on the table.
Chris Olmsted Tue 17 Jun 2014
I'm in the individual tag camp and here's why:
I believe that drilling down to more specific points should be as simple as adding additional tags to a query. There will be exceptions, like blowDown or makeUp, where a concatenated approach is necessary because the individual tags are meaningless (blow and down). Individual tags will make haystack easily expandable and flexible. If every tag combination has a unique concatenated name, like primaryChilledWaterFlowMeter, then we might as well use point names from the control system.
As for the potential for exclusionary boolean algebra: why not add more tags? From the example above:
I'm not sure what a regular water meter is, but let's say it's a domestic meter, then you can use:
I believe anytime we have a case where a set of points can only be accessed though exclusionary logic, that is a sign that we need to add tags to the library.
Brian Frank Wed 18 Jun 2014
Okay, if everyone is willing to make a breaking change then my vote is to go individual tag route. That is consistent with what we've done with points
dischageTemp
went todischarge air temp sensor
.Going thru previous posts, this what I have for full list of meter types:
Keith Bishop Wed 18 Jun 2014
I think the individual tag model is where we should go. I do think we should discuss a transition plan for this change.
If we are considering splitting out combined meter tagging, should we also look at the other combined tags and discuss their separation.
As far as the meter list above, I don't know that "flow" is a tag that we want to adopt on these fluid meters. Fluid metering is measuring flow. With the water meter, should we use the tag "domestic" instead? The steam "flow" tag should be able to be dropped. Again, a meter on a steam system is going to measure flow. There shouldn't be a need to differentiate it from other sensor type objects.
The "steam" tag should be able to be dropped from the condensate meter. The "condensate" tag should indicate a piece of a system that represents a portion where a phase change has happened from the steam portion of the system. I understand that in many cases condensate is measure instead of steam flow to more accurately measure the mass flow rate of the system to avoid the inherent density compensation issues with measuring steam. No matter the underlying intent, it is still a condensate meter.
I don't know that we have a good use case at this time but as this model grows I could see the need for other metering types for things such as pressurized air systems or specialized water systems such as deionized water. With that said, I think the current list represents the majority of employed meters.
One last thing. With this discussion about metering I think we should constrain the use of "meter" to objects that either measure a rate of use or a totalized consumption.
John Sublett Wed 18 Jun 2014
I like the individual tag approach. It makes for much flexible queries and I think it should be used consistently throughout the model.
If you are considering a breaking change like breaking "elec" apart from "meter", how about fixing the energyRate and energyConsumption tags to be consistent too and removing the kw and kwh tags? Units should be specified in the "unit" tag.
In fact, in the interest of individual tags, shouldn't "flow", "energy", "rate", and "consumption" all be individual tags?
Keith Bishop Wed 18 Jun 2014
Can we assume that metered energy points are "rate" points unless a hisTotalized tag is used? Or should we use the unit tag to make this differentiation (kW or gpm are rates whereas kWh and gal are consumption)?
Richard McElhinney Wed 18 Jun 2014
@Brian, I'm on board with the individual tag approach as well. It is definitely more consistent with the other models we've got. So +1 for me.
In terms of other meters, now that we've (basically) agreed on the individual tag approach it should be easier to capture new meters so perhaps it's worthwhile moving forward with what you have proposed.
WRT what @John Sublett says, he raises a good point in proposing to break up other compound tags. While we are making a breaking change here is it better to "rip the bandaid off quickly" and review the rest of the current tags to see if there is areas where we can encourage consistency.
This might be too much of a breaking change to bite off in one go, but just throwing it out there for discussion.
For example a quick look at the documentation reveals some potential for further consistency:
I guess this kind of hijacks the thread and maybe this should be broken out into another topic so it doesn't get confused with metering conversation.
However, I would be interested in feedback.
Zach Schweinfurth Fri 20 Jun 2014
After reading through this thread, I want to express my support for the individual tag approach. I think Chris Olmsted is correct in saying that if we build up unique compound tags for many things, then we should just use point names from control systems. Moreover, I agree with Chris that in some cases, compound tags might be necessary when an atomic tag means nothing really by itself. I don’t think we should be so rigid with the atomic tags that we don’t add compound tags in when necessary.
A general guiding rule(s) could be something like: Tags should contain no more words than necessary to identify the underlying concept. And if a compound concept can be equally conceptually represented by two (or n) atomic tags, then the atomic tags should be used over the compound tag provided that each atomic tag represents an engineering concept of importance on its own.
So taking the electric meter example, electricity and meters are conceptually separate things that have engineering significance on their own. And when combined, the tags create a concept distinct from their parts, which is also of significance to engineering.