Estimated Flow Rate

Based on the conditions you provided, you can't.

More explanation would be helpful. What conditions?

Your pressure measurement is downstream of a pressure regulator which controls the downstream pressure to its setpoint pressure.

Pressure regulators are proportional and exhibit 'droop', a drop off from the control setpoint, as the load increases.

The phenomenon you are seeing is the droop in pressure with the increased load of the 2nd appliance.

I suppose you could record the pressure droops at different consumption rates (since you know an appliance is Xcfm), then create a look-up table to reference what the consumption is at a given droop pressure.

Thanks for you response, BUD. I understand that the increased load drops the pressure in the meter. I have done this in the lab using a simple piping network, I would like to extrapolate to a full house.

I am aware there is a pressure regulator before the meter that provides the meter with a constant pressure. The changes after that are caused in the piping network (runs, fittings, valves and what not.).

Are you telling me that even if I know the pressure drops in every network, I cant determine the combined effects on the flow rates??

please advise

Why on earth would one want to (rhetorical question - NNTR)?

Get the sytem installed by a qualified Gas Fitter, who will be CORGI-registered in the UK for example, and get a signed completion certificate to present to the company responsible for providing fire insurance cover for the premises. Anything less is fraught with unneccessary risk and really only for those applying for a Darwin Award.

He's only writing a paper, as far as we can tell he's not threatening to install anything.

Well, that's a relief.

Arooooooooooogah! Undefined "we" alarm.

Ha ha ha.

And the Gas Safe Register superseded CORGI several years ago.

"provides the meter with a constant pressure"

That is a 'misunderstanding' of the operation of mechanical proportional regulators.

The graph below shows a typical 'constant' pressure regulator "pressure vs flow" curve.



Your first appliance operates at pressure A. Both appliances operate at pressure B where the flow is doubled. It's the nature of the beast.

Pressure drops in piping are not constant; pressure drops vary with flow rate, so "know(ing) the pressure drops in every network" becomes a significant measurement issue.

From your figure, I see that the pressure drop in the piping network is not important.

It seems the droop and the established set point can be used to estimate the problem I originally described. If I have one data-point, I can fit the flow on line given by your curve.

Is this reasonable?

I am not designing anything, I am writing a paper of gas load estimates, and I have a simulation of gas consumption by different appliances. I wanted to include the reduced flow rate we get when we sum two appliances into the model, and I wanted to be able to justify the reduction factor, instead of pulling it out of nowhere.

Also, do you have any idea what the maximum variations expected?

Keep in mind that a properly sized and fully functional pressure regulator will stabilize the downstream pressure within very close tolerance.

There will be a brief drop in pressure when the flow demand suddenly increases but the regulator should react and control the pressure within the calibration tolerance range.

This pressure drop will be minimal unless the pressure regulator is not of the correct type, is undersized, not calibrated properly, or not functioning correctly.

You know the pipe size, the system dynamic losses (as stated/adopted), and the total appliance flow demand per household so you should be able to apply the nominal useage derating factor, multiply by the number of households attached to the distribution system then calculate the total flow demand on the system.

The droop curve is representative, it is not for any specific brand or model regulator. My marks are arbitrary, and have no value associated with them other than an illustration of droop. Each model regulator has its own 'curve'.

Droop curves are available from high end industrial regulator vendors, but I don't know if you can weasel that information out of the gas company.

On the other hand, you can measure the pressure (and its droop) with a water manometer with a length of plastic tubing and some tap water. Then you can see if you can find a fuel DP vs Heat curve like the one below for an industrial atmospheric (not forced combustion air) burner burners.

With such a graph, you can read directly what a drop in fuel supply pressure will do to the BTU output (assuming that the entire DP is across the burner nozzle)."Appliance" manufacturers may or may not publish similar data.

My impression is that proportional regulators work 'good enough'. The output might not be 'constant' but it remains close enough to setpoint for typical residential service. Industrial service is another matter.

Interestingly enough, I did measure the pressure drop vs. flow rate curve for the specific meter I am talking about using a manometer. Here is my setup and data.

I measured it accross the meter, meaning the inlet and outlet. The piping network is also shown in the second picture, but not much loss there. I am still unclear as to how I can use this data to get the information I need.

Do I use the pressure drop in the network and scale the flow by this curve.

You insight in the topic is very helpful.

Interesting...your curve is nearly linear, except near the left end. Usually it is considered that ΔP is proportional to Q² (or, equivalently, Q proportional to √ΔP).

To add to what others have said - the actual pressure downstream of the regulator falls as the flow increases. The regulator characteristic should show a number of curves of actual pressure vs flow, for a range of set pressures. Naturally the drop-off in pressure for a given flow is lower the bigger the regulator.

If the appliances do not have an additional pressure regulator the flow taken depends on the pressure. You need to know the pressure/flow characteristic of the appliances, and take account of the pipe pressure loss. You might need an iterated calculation if you want to be accurate, as pressure and flow are interdependent.

Also not sure whether "at the same location, meaning similar pipe configuration" means both appliances are fed from a common pipe, or each from a similar pipe configuration from the meter.

But having said that, if the appliance is say a boiler it probably has its own pressure regulator, as the pressure needed at the burner is a good bit below supply pressure. The boiler regulator needs a minimum input pressure to provide correct output pressure, but as long as pipe friction plus main regulator drop do not cause pressure below that, flow is not affected. That's why the boiler regulator is there, to give some independence from supply pressure.

Are you writing this paper for mcodesmart ?

If yes, that puts all this in a different light, at least for me.

Do you get paid for these? If yes, full disclosure is in order.

That's a no. And what is that anyway?

I am writing an academic paper, not a blog.

It is your home page.

So, do you have any constructive suggestions?

Yes, give credit to everyone here who contributed to your academic paper.

Ok. Thanks. I meant regarding the discussion

No, I don't.

You seem to be in very good hands with Bud.

>When you turn one on, you see 50 CFH in the meter. When you turn the other one on, you don't see (50+50)CFH but a slight reduced value.

> the reduced flow rate we get when we sum two appliances into the model

> measured pressure drop across the meter

I misinterpreted the word 'value' in your original post as 'pressure', not flow. I thought your pressure dropped slightly when the 2nd appliance came on-line.

But as I re-read, it appears that you are saying that the flow meter does not see 100 CFH when two 50 CFH appliances are turned on, one after the other. Did both appliances function normally? Good blue flame?

The meter in the photo appears to be a conventional diaphragm meter, a positive displacement technology with a typical 1% accuracy. Even the low end residential meters are typically capable of 250 CFH at 0.5"drop.

Your photo does not show the regulator. Is it a typical gas company regulator? Upstream of the meter?

My comments on droop are not applicable to the issue of a meter doesn't indicate increasing flow with an increased load. Ordinary regulator droop at 100CFH will not cause a meter to erroneously indicate a decreasing flow with an increase in the downstream load.

What's the history on the meter?

The meter is a new american standard meter. I got this result in the lab, I have a bunch of valvs that simulate the aplinaces.. and when I combine them, I do not get them to sum up to the independent values. I figured friction would be the reason, but I still do not have a grasp of this?

And I also do not know if this is something that happens in a residential setting?

Take a typical house:

Stove, oven, gas water heater, gas furnace, gas dryer. The expensive houses might have a gas grill and gas fireplace.

Each appliance has it's own regulator except maybe, the gas dryer. PS: They hide.

Every house has it's own friction losses. Pipe size, pipe length, elbows, etc.

Uh-oh; now you're KeepinItComplicatedSmartie!

Agreed, see #10

You are assuming that non of the appliances have their own internal pressure regulator set slightly below the main regulator pressure, which in the UK at least, is usually the case. Flow restriction from these secondary regulators would allow the pipe network pressure to rebuild to the main regulator pressure. If this does not happen then the pipe diameters are insufficient for the appliance flow rates. If you are not seeing this in your lab setup, is your setup correctly simulating the conditions present in a house?

Agreed, see #10

Number of moles (ni) = number of moles out (no)

Volume in = Volume out (unless there is a change in temperature)

PV=nRT &

P1V1=P2V2 or in your case PtVt=P1V1+P2V2+...PnVn (assumes constant T)

By knowing the properties of the pumps power and rotating velocity, you can calculate the mass flow rate which is pumped to the pipes. Then by the following formula the velocity can be calculated:

Mass flow = Density * Velocity * Pipe Area