Proving RMFDs with ATC for the ATC Steering Committee

1
Proving RMFDs with ATCfor the ATC Steering
Committee

• Ross Andersen
• New York State
• August 27-29, 2007

2
Appreciation to Measurement Canada

• Invitation from
• Alan Johnston
• Gilles Venet

• Field support from
• Dennis Beattie
• Bob Perrin
• Steve Kotack
• Rod Maxwell

Thanks to several gas stations that also
cooperated.
3
Objectives

• Observe necessary changes to RMFDs to support
  ATC (slideshow)
• Touch on Test Equipment needs
• Explain operation of mechanical and electronic
  ATC devices
• Describe HB44 approach to proving ATC
• Relate experience and results proving ATC devices
  using HB44 approach
• Explain the Canadian approach to proving

4
Important Caveat

• A Retail Motor Fuel Dispenser will be more
  accurate than our ability to test it!
• Much of the variation we see in official testing
  is not due to variations in the performance of
  the device!
• High compliance (95) is an indication

5
Differences with ATC

• Pump changes
• Thermometer Probe with sealing
• Thermometer Well
• Test Mode each manufacturer unique

• Equipment Needs
• Accurate and reliable thermometer
• VCF tables for products dispensed

6
Thermometer used in Canada

• Specifications
• Reliable thermistor technology
• Intrinsically Safe
• Reading to 0.1C or 0.1F
• Accuracy to 0.2C or 0.3F
• Cost 200 60 for suitable flexible probe

Other brands and models may also meet our needs.
7
Mechanical ATC

• Sensor usually a bellows that mechanically
  adjusts the transmission
• Must have provision to deactivate for gross
  proving ratio11
• Cant get both gross and net indication from same
  delivery, except old-style stacked registers

8
Electronic ATC

• Given k calibration factor (gross pulses per
  gallon)
• Gross Volume n / k
• Net Volume SVCF / k

9
HB44 Proving Approach

• Evaluate Gross Delivery Error, i.e. gross prover
  volume gross meter indication
• Evaluate Net Delivery Error, i.e. net prover
  volume net meter indication
• Net prover volume calculated as gross prover
  volume x VCF for product temperature in prover
  and specified product density
• Mechanical systems need two test drafts for gross
  and net while electronic only needs one draft
• Evaluate agreement between Gross and Net

10
HB44 Proving Approach

• Works for mechanical and electronic ATC
• Agreement requires ATC device to have accurate
  temperature sensor
• Proving requires accurate temperature
  measurements at meter and in prover
• Best if temperature in equilibrium
• Requires some calculations to be fully
  comprehensive

11
ATC Testing August 8 PM

• Testing at Esso station in Ottawa
• Gilbarco B78 series dispensers
• Temp differential (ambient/product) 5C
• Evaluate temp changes in thermo well using
  readings at 5,10,15 and 20 l
• Getting stable temperature in prover
• Evaluating gross/net agreement
• Calculations on all drafts (inc. wet down)

12
ATC Testing August 9 PM

• Testing at Sunoco station in Ottawa
• Wayne 21V series dispensers
• Looking at repeatability of results
• Temp at well at 6.6 and 13.3 l (1/3 2/3)
• Evaluating LIG cup thermometer in prover
• Evaluating gross/net agreement
• Calculations on all drafts (inc. wet down)

13
Temperatures in Thermo Well
14
Delivery Errors (prover meter)

• Raw Gross Prover reading w/o correction for
  product temp change M to P or prover temp
• Raw Net Calculate as raw gross x VCF for 730
  kg/m3 at prover temperature
• Corrected Gross Prover gross corrected for
  product temp change M to P and prover temp
• Corrected Net Calculate as prover gross (with
  correction for prover expansion) x VCF for 730
  kg/m3 at prover temperature

15
Overall Results 8/8 and 8/9
16
Observations and Conclusions

• Gross/Net Agreement (Corrected)
• Most tests passed gross/net agreement
• Obtained good agreement even with LIG cup
  thermometer
• Only 2 wet down drafts (8/8 at 1448 and 1516)
  with large temperature changes failed to conform

17
Overall Results 8/8 and 8/9
8/8 8/9
18
Observations and Conclusions

• Ambient temperatures impact results
• Ambient 5C hotter than product on PM 8/8, all
  corrected gross results lower than raw (prover
  temp always warmer than average product)
• Ambient product on AM 8/9, all corrected gross
  results higher than raw (prover temp always
  cooler than average product)
• Saw 20 ml impact on results 0.1
• May be more if differential larger!

19
Overall Results 8/8 and 8/9
20
Observations and Conclusions

• Raw Values - Gross vs Net
• Excellent agreement
• Only two results (8/9 1021and 1023) failed to
  meet requirement used Liquid in Glass cup
  thermometer in prover for those tests on pump 1
  94 octane

21
Corrected Net vs Raw Net
22
Observations and Conclusions

• Comparing Corrected and Raw Net
• Generally within 5-7 ml 0.035
• Difference mostly in the prover correction for
  temperature 23 C
• May need to do extensive calculations to get good
  results only at extreme temperatures!
• Can run additional tests to get better
  temperature equilibrium

23
Canadian Proving Approach

• With equilibrium we assume
• Net Gross x virtually constant VCF
• Calculate VCF used by the dispenser
• Pump VCF Net / Gross
• Look up equivalent product temperature from VCF
  table
• Equivalent product temperature must agree with
  observed meter temperature in thermo well within
  1 C / 1.8 F

24
Canadian Proving Approach

• Example 8/8 at 1510
• Meter net / Meter gross 0.98977
• Equivalent temperature from VCF table for 730
  kg/m3 23.2 C
• Observed temperature at thermometer well 23.5 C
• Agreement 0.3 C

25
Canadian 1C Agreement
8/8 PM 8/9
AM
26
Observations and Conclusions

• Evaluating the Canadian Approach
• Requires temperature equilibrium
• Simple requires only a hand calculator to
  divide net by gross and VCF tables to look up
  temperature for that calculated VCF
• Lag and ambient heating in thermo well impacts
  results
• Can run additional product to get better
  temperature equilibrium
• Does essentially same thing as HB44

27
Observations and Conclusions

• Proving with ATC will not be a significant
  undertaking! We can do this easily!
• While bottom-drop provers with reservoirs will be
  ideal, they are not a necessity
• If using hand poured measures it is vital to
  ensure reasonable temperature stability
• Wont require use of computers in the field, as
  we can create simple aids to assist in making
  important corrections