Accuracy of Single-Jet and Multi-Jet Water Meters under ... · Pressure sensor P1 Hydrostatic...

KIT – The Research University in the Helmholtz Association

KARLSRUHE INSTITUTE OF TECHNOLOGY, INSTITUTE FOR WATER AND

RIVER BASIN MANAGEMENT

www.kit.edu

Accuracy of Single-Jet and Multi-Jet Water Meters under the

Influence of the Filling Process in Intermittently Operated Pipe

Networks

IWA Water IDEAS 2016 Conference, 21.10.2016, Bologna, Italy

David Walter, Miran Mastaller and Philipp Klingel

Karlsruhe Institute of Technology

Institute for Water and River Basin Management

2

Purpose and Function of Water Meters

Purpose

Economic base of water supply

Determination of delivery to customers

Determination of water balance and water losses

Accuracy of Single-Jet and Multi-Jet Water Meters under the Influence of the Filling

Process in Intermittently Operated Pipe Networks

21.10.2016

Function of single-jet and multi-jet meters

Impeller set in rotational motion by single or multiple jets

Rotational velocity proportional to flow velocity

Number of revolutions proportional to cumulative volume

Low-flow indicator can be used for determining flow over time

Multi-jet meters use bypasses for calibration

Multi-Jet Meter Single-Jet Meter

Multi-Jet Meter

Single-Jet Meter

Low-flow indicator



3

Accuracy of Water Meters in Intermittent Supply

Accuracy of water meters

Accuracy requirements are regulated in ISO 4064-1: 2014

Meters have to be used in line with their conceptual design

E.g. the meter must be completely filled with water and vented



21.10.2016

Intermittently operated water distribution systems

Pipes are not constantly filled with water and under pressure

Intermediate storage of water in private tanks

In areas with high water scarcity and flat-rate tariffs:

Service connections constantly opened to access water

Impacts on network and meters

Pipes drained after each supply period

Air exits network in subsequent filling process

Part of the air flows through service connections

Meters being used contrary to their initial design

Water

Air

Filling Process

of Pipe Network



5

Accuracy of Water Meters in Intermittent Supply



21.10.2016

Influences on measurement accuracy

Air flow before the arrival of the water front

Impact of the water front on the impeller

Air intrusion in the front part of the water column

Unsteady flow conditions and inertia forces affecting on meter mechanics

Common inaccuracies during steady-state flow (restricted by ISO 4064-1: 2014)



6

Approach

Investigation of water meter accuracy under influence of pipe filling process

Horizontally mounted single-jet (Q3 = 2.5 R80H) and multi-jet (Q3 = 4.0 R80H) meters

Determination of correlation of measurement error, pipe pressure and air volume

Experimental Parameters:

Pipe pressure p1,stat

Volume of air in front of the meter Vair / pipe length LS

Dry and wet meter casings

p1,stat Vair

wet meter casing

wet meter casing

dry meter casing

dry meter casing LS

p1,stat 0.1 – 1.0 bar

Vair 0.5 – 7.8 L

LS 1 – 25 m



21.10.2016



7

Experimental Test Set-Up

Pressure sensor P1

Hydrostatic pressure p1,stat

Start of water column

movement at pressure drop

(tp1,start)

Pressure sensor P2

Arrival of water front

at meter at pressure

rise (tp2,start)

Pressure sensor P3

Determination of real

volume (water) Vt

Optical sensor I

Determination of total

volume (water + air) Vwm

Determination of meter’s

flow over time Qwm(t)



21.10.2016



8

Experimental Test Set-Up

Pressure sensor P1

Hydrostatic pressure p1,stat

Start of water column

movement at pressure drop

(tp1,start)

Pressure sensor P2

Arrival of water front

at meter at pressure

rise (tp2,start)

Pressure sensor P3

Determination of real

volume (water) Vt

Optical sensor I

Determination of total

volume (water + air) Vwm

Determination of meter’s

flow over time Qwm(t)



21.10.2016



9

Determination of Measurement Errors



21.10.2016

Etotal total error during filling process

Eair error caused by air flow

Efront error caused by impact of water front on impeller

Eintr error caused by air intrusion in front part of water column

Eunsteady error caused by unsteady flow and inertia forces of meter‘s mechanics

Esteady common meter errors of during steady-state flow (ISO 4064-1: 2014)

no precise differentiation with described test set-up



10




21.10.2016

Etotal

Eair

Efront

Eintr

Eunsteady

Esteady

Erest



11




21.10.2016

Etotal 𝐸𝑡𝑜𝑡𝑎𝑙 = 𝑉𝑤𝑚 − 𝑉𝑡 − 𝑉𝑤𝑎𝑡𝑒𝑟 total error during filling process

Eair error caused by air flow

Erest 𝐸𝑟𝑒𝑠𝑡 = 𝐸𝑡𝑜𝑡𝑎𝑙 − 𝐸𝑎𝑖𝑟 remaining error (Efront + Eintr + Eunsteady + Esteady)

𝐸𝑎𝑖𝑟 = 𝑄𝑤𝑚 𝑡𝑡𝑝2,𝑠𝑡𝑎𝑟𝑡

𝑡𝑝1,𝑠𝑡𝑎𝑟𝑡

𝑑𝑡



12

Measurement Error Erest



21.10.2016

Very low positive and negative values near zero

No dependence of Vair or p1,stat

Erest can be neglected for estimating Etotal

𝐸𝑟𝑒𝑠𝑡 ≈ 0

𝐸𝑡𝑜𝑡𝑎𝑙 ≈ 𝐸𝑎𝑖𝑟



13

Measurement Error Eair for Single-Jet Meters

and Dry Meter Casing



21.10.2016

Eair is independent of p1,stat for p1,stat ≥ 0.2 bar

Linear correlation between Eair and Vair for Vair ≥ 1.08 L

Non-Linear correlation between Eair and Vair for Vair < 1.08 L

Eair is dependent of p1,stat for p1,stat = 0.1 bar due to unstable build-up of water front



14





21.10.2016

Eair is independent of p1,stat for p1,stat ≥ 0.2 bar

Linear correlation between Eair and Vair for Vair ≥ 1.08 L

Non-Linear correlation between Eair and Vair for Vair < 1.08 L

Eair is dependent of p1,stat for p1,stat = 0.1 bar due to unstable build-up of water front



15





21.10.2016

Eair,mv,d as mean value error curve

Gradient and deviation in volume measurement of 96% in linear range

Constant vertical offset of ΔEinert in linear range due to starting resistance of impeller

Decline of ΔEinert in non-linear range due to compression of air in front of the water column



16





21.10.2016

Eair,mv,d as mean value error curve

Gradient and deviation in volume measurement of 96% in linear range

Constant vertical offset of ΔEinert in linear range due to starting resistance of impeller

Decline of ΔEinert in non-linear range due to compression of air in front of the water column



17





21.10.2016

Estimation of Etotal

𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 = 0.93 ∙ 𝑉𝑎𝑖𝑟 − 0.56 L + 𝐸𝑟𝑒𝑠𝑡 for 𝑉𝑎𝑖𝑟 ≥ 1.08 L (1)

𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 = 0.4 ∙ 𝑉𝑎𝑖𝑟2 − 0.04 ∙ 𝑉𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 for 𝑉𝑎𝑖𝑟 < 1.08 L (2)



18


and Wet Meter Casing



21.10.2016

Significantly smaller values of Eair for p1,stat = 0.1 bar water remains in meter casing

Abruptly larger values of Eair for p1,stat = 0.3 bar water being pushed out of meter casing

Fluctuating values of Eair for p1,stat = 0.2 bar unstable condition

Mean value error curve can be divided in an upper and a lower range



19


and Wet Meter Casing



21.10.2016


𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 = 0.76 ∙ 𝑉𝑎𝑖𝑟 − 0.63 L + 𝐸𝑟𝑒𝑠𝑡 for 𝑉𝑎𝑖𝑟 ≥ 1.08 L; 𝑝1,𝑠𝑡𝑎𝑡 ≥ 0.3 bar (3)

𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 = 0.19 ∙ 𝑉𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 for 𝑉𝑎𝑖𝑟 < 1.08 L; 𝑝1,𝑠𝑡𝑎𝑡 ≥ 0.3 bar (4)

𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 = 0.11 ∙ 𝑉𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 for 𝑝1,𝑠𝑡𝑎𝑡 = 0.1 bar (5)



20

Measurement Error Eair for Multi-Jet Meters and

Dry Meter Casing



21.10.2016

Eair is dependent of p1,stat and becomes smaller with higher values of p1,stat for p1,stat ≥ 0.2 bar

Decrease of Eair for p1,stat = 0.1 bar due to unstable build-up of water front

Separation of error curves in linear and non-linear section between Vair = 1.74 L

Gradient in linear range of 58% for 1.0 bar to 150% for 0.2 bar

Dependency of Eair on p1,stat may caused by vibrations of the impeller and existence of bypass



21


Dry Meter Casing



21.10.2016


𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡

= 1.73 − 1.15 ∙ 𝑝1,𝑠𝑡𝑎𝑡 ∙ 𝑉𝑎𝑖𝑟 − 2.05 − 1.5 ∙ 𝑝1,𝑠𝑡𝑎𝑡 + 𝐸𝑟𝑒𝑠𝑡 for 𝑉𝑎𝑖𝑟 ≥ 1.74 L (6)

𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡

= 0.24 − 0.1 ∙ 𝑝1,𝑠𝑡𝑎𝑡 ∙ 𝑉𝑎𝑖𝑟2 + (0.15 − 0.13 ∙ 𝑝1,𝑠𝑡𝑎𝑡) ∙ 𝑉𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 for 𝑉𝑎𝑖𝑟 < 1.74 L

(7)

p1,stat in bar



22


Wet Meter Casing



21.10.2016

Significantly smaller values of Eair for all pressures water remains in meter casing

Velocity of air flow not high enough to lift initial water out of the lower chamber


𝐸𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 = 0.15 ∙ 𝑉𝑎𝑖𝑟 + 𝐸𝑟𝑒𝑠𝑡 (8)



23

Conclusion



21.10.2016

1. Single-jet and multi-jet water meters, dry and wet meter casing:

Measurement error Erest consists of very low values that fluctuate around zero

Erest can be neglected for estimating the total error Etotal

2. Single-jet water meters, dry meter casing:

The error is independent of pipe pressure p1,stat and only depends on the air volume Vair

Estimation of total error via equation (1) and (2)

3. Single-jet water meters, wet meter casing:

Additional dependence of error on pressure p1,stat for small pressures caused by increase of

the impeller’s starting resistance by initial water

Estimation of total error via equation (3), (4) and (5)

4. Multi-jet water meters, dry meter casing:

The error is dependent of pipe pressure p1,stat since the impeller starts vibrating and the air

flow shifts towards the bypass

Estimation of total error via equation (6) and (7)

5. Multi-jet water meters, wet meter casing:

The error is independent of pipe pressure p1,stat since the initial water remains in the chamber

and affects the rotational motion of the impeller for all tested pressures

Estimation of total error via equation (8)

Accuracy of Single-Jet and Multi-Jet Water Meters under ... · Pressure sensor P1 Hydrostatic...

Documents

Transcript of Accuracy of Single-Jet and Multi-Jet Water Meters under ... · Pressure sensor P1 Hydrostatic...