The Importance of Measurement in Determining Gas Emissions ... gas... · The Importance of Measurement in Determining Gas Emissions ... Determining Gas ... Low gas price = less incentive

The Importance of Measurement in

Determining Gas Emissions from

Flares and Vents

Technical Consultant

Dr. Jeff Gibson

Presented by

Dr. Norman Glen

12 June 2013

Contents

Introductory slides:

Why is measurement important?

The UK National Measurement Office (NMO) and the National

Measurement System (NMS)

Main presentation:

Emissions from flares / Worldwide legislation

Methods used to determine quantity flared

Issues affecting flare measurement

Case study – The use of CFD simulation to determine the

installation error in flare gas meters

Any questions?

Measurement plays a fundamental part in the innovation

process. To develop new products and processes, companies

need to measure quantity, quality and performance.

To trade successfully, companies utilise a regulatory

framework, based upon measurement confidence, ensuring

access to global markets that are fair and open and without

unnecessary barriers to trade.

National Measurement Office (NMO)

Why is measurement important?

National Measurement System (NMS)

The NMS is a Network of laboratories and processes that

provide measurement standards and calibration testing

facilities.

It maintains the national measurement infrastructure

It represents the position of UK measurement

Internationally and influences the development of

standards.

Energy,

Oil & Gas

Low Carbon

TechnologiesEnvironment

UK National Standard for Fluid Flow Measurement:

A Global centre of excellence for flow measurement and

fluid flow systems serving both industry and Government

World class research, development, modelling, calibration,

consultancy, measurement and testing services to clients

across many sectors including: Energy, Oil & Gas, Low

Carbon Technology and the Environment

NEL (Scotland, UK)

Measurement system requirements

Design of system will depend on the “type” or

purpose of metering required:

Custody transfer

Allocation

Management and Control

Process operation

Environmental reporting (e.g. EU ETS)

The cost of errors

Oil field producing

20,000 BBL/Day

Cost per barrel: $100

Revenue: $2.0M per day

Suppose the meter under-reads (error) by

1%

3%

= Loss of revenue $20,000 per day

= Loss of revenue $60,000 per day

Main Presentation:

The Importance of Measurement in

Determining Gas Emissions from

Flares and Vents

Main Purpose of Flare System

Emergency relief system (fire risk, prevent over-

pressurisation, leaks and spills)

To allow for depressurisation for routine shutdowns and

maintenance periods

To vent-down vessels and pipework during equipment

trips and “black-outs”

Purging of O2 prevents pre-ignition

Steam/air injection reduces smoke and improves

efficiency

Yes, require an economic route for the gas, e.g.

Sell as gas, burn as fuel, LNG export, distil as

chemical product (methanol, dimethylether...)

Pipeline or LNG plant = large investment

Low gas price = less incentive to capture gas

Flare Gas Recovery (FGR) system is needed to allow

capture with safe operation

Can’t the gas be reused instead

of flaring it?

Flare normally

extinguished

FOV lifts when

pressure exceeds

safe limit

Global Gas Flaring/Venting Impact

Over 150 billion m3 gas flared/vented per year

= 30% of entire EU gas usage per year

= 30.6 billion USD in revenue wasted

= 360 million tonne CO2 equivalent (tCO2e)

Source: World Bank Global Gas Flaring Reduction (GGFR)

Partnership

Country

Quantity

Flared

(bcm/year)

% of Global

Flaring

(by volume)

1 Russia 37.4 26.72 Nigeria 14.6 10.43 Iran 11.4 8.14 Iraq 9.4 6.75 USA 7.1 5.16 Algeria 5 3.67 Kazakhstan 4.7 3.48 Angola 4.1 2.99 Saudi Arabia 3.7 2.6

10 Venezuela 3.5 2.5

Total 100.9 70

(Sources: Worldbank, 2011 data)

Top 10 Gas Flaring Nations

76% of

associated gas

40% of

associated gas

NOAA Satellite Imaging Datahttp://www.ngdc.noaa.gov/dmsp/interest/gas_flares.html

Gulf of MexicoNorth Sea

Nigeria

Flare Header lines

Liquid knockout / seal drums

Flare gas recovery (FGR) system

Pilot gas ignition system

Steam or air injection system

Elevated stacks or ground-based

flare-battery (onshore only)

Incinerators may be used

Key Elements of a Flare System

Ground FlaresElevated Flare Stacks

Emissions from Flaring & Venting

Greenhouse gases (GHGs): H2O, CO2, CH4, N2O, O3,

SF6 (electrical industry), Fluorocarbons and

Hydrofluorocarbons (refrigerants)

Main GHGs emitted: CO2 & CH4

Other unwanted gases (and effects):

SO2 & NOx - acid rain, ozone depletion

Non-methane VOCs – alkanes, alkenes, aromatics...

CO, particulates, smoke (incomplete combustion)

H2S, mercaptans (highly toxic, acidic, odours)

Noise & light pollution, radiation hazards…etc.

Legislation

Just about all nations have some form of legislation

for flaring and venting

Some more rigorously enforced than others (e.g.

Nigeria vs European Union)

Most prominent schemes: EU (ETS), Norway (ETS,

Taxation), Canada (e.g. ERCB Directive 060) and

USA (e.g. Rule 30 CFR 250, Subpart K)

Backed-up with Measurement & Reporting

Regulations (e.g. EU ETS MRR from Jan 2013)

EU Emissions Trading Scheme (1)

EU Kyoto target is combined 20% reduction by 2020

EU ETS is seen as key policy instrument for

Controlling and Reducing GHG emissions (others:

renewable, energy efficiency & CCS)

EU ETS is predominantly a CO2 Cap-and-trade

Scheme

Targets largest emitters (> 20 MWth aggregate)

Scheme commenced 2005, now in Phase III

Measurement uncertainty

Requirements set by tier level depending on total CO2

Should take into account stated uncertainty in equipment

(and/or calibration) and in use

Data, and associated measurement uncertainty, must be

verified by an accredited body

Traceability of equipment

Regular calibration, adjustment and checks (ISO 17025

lab or equivalent)

Measurement equipment traceable to International

Standards

Where no traceable standards available - identify other

alternative control and verification measures

EU ETS Phase III Monitoring &

Reporting Regs (MRR)

Volume

flow, Q

Steam/Air

Acid gas

(CO2/H2S)

Q = Metered or Calculated

EF = From gas composition

OF = Typically a constant value

Open flame

Air dilution

(unknown)

Mass CO2 =

f(Q, EF, OF)

EU ETS “Calculation Method”

Emission

“Cloud”

where:

Q = Standard volume (m3)

EF = Emission Factor (tCO2/m3)

OF =Oxidation Factor (-)

CO2 Emissions = Activity Data x Emission Factor x Oxidation Factor

tCO2 = Q x EF x OF

• Voume, Q (m3):

(Amount of Gas Flared)

• Emission Factor, EF(tCO2/m3):

(Process modelling software)

• Oxidation Factor, OF (-):

(Assume = 1.0, all C CO2

UK Benchmark = 0.98)

Measurement Uncertainty:

Cat A Cat B Cat C

< 50 ktCO2e 50-500 ktCO2e > 500 ktCO2e

± 17.5% ± 12.5% ± 7.5%

Uncertainty given in EU ETS MRR

EU ETS Minimum Tier Requirements

Very wide velocity range (> 1000:1)

Minimal pressure drop required

Large line sizes

Meters often welded (“hot-tapped”) to existing pipe

Low flow (< 0.1 m/s):

• Resolution problems

• Re < 10,000 = Transition, Stratification,

High instability

Blow-downs (50 - 100 m/s+):

• Compressible, unstable and noisy

• Rapid temperature drops

• Liquid carry-over/drop-out, suspended solids

Issues for Flare Measurement

Calibration & Traceability Issues

Difficult to remove meters for lab calibration (full flow range

unlikely to be achievable in any case)

Calibration unlikely to be done except prior to commissioning

spool-pieced meters

In situ verification flow tests:

Gas tracer techniques (e.g. Kr85, SF6, He…)

All in situ methods “at mercy” of prevailing flow conditions

/ unknown leaks / inadequate tracer mixing….

Insertion probes - H&S issues?

Comparison against other meters (e.g. Deliberately divert

gas to flare through a “reference” meter = costly)

Basic in situ checks (Note: not flow tests)

• Post-installation verification – dimensional checks

• Zero-checks of electronics/cabling

• Online „self-check‟: e.g. Check speed of sound is

“sensible” (ultrasonic meters)

• Extraction/checking of transducers “in a box”

containing reference gas (e.g. nitrogen) against a

baseline reading

Checks transducers/cabling/electronics, but not

response under flowing conditions

Verification Checks

Many technologies available as insertion-type (or

separate transducers) that can be installed through

bosses. Here are some of these:

Insertion turbine

Insertion vortex

Averaging pitot

Thermal mass

Ultrasonic meters

Optical correlation meters

Limited range

Wider range

Dedicated Flare Meters

These are “Point” (or at best “Line”) measurement

devices

Only “see” only a portion of the flow and measure local

velocity (need corrections to get to average velocity

and volume flow)

Multiple paths / points can improve measurement, but

low flow region remains a challenge

Cannot be calibrated unless pre-installed in a spool-

piece that can be removed for periodic flow calibration

Insertion-type meters

Use transit-time method to get velocity (volume via

correction and geometry):

Many applications worldwide both on and offshore

Very wide flow range with low overall DP

Density (possibly EF) from MW = f(SOS, T, p….)

Single-path measurement (affected by flow profile)

Large errors/scatter at low flow

“Beam Drift” at high flow

Liquids cause over-reading and SNR problems

Ultrasonic Flare Gas Meters

When can indirect (estimation) methods be used to

determine flare quantities?

de-minimus quantities (< 2% of total emissions)

When installing meters is unfeasible

For periods where meters malfunction (back-up)

“Un-measurable” conditions (very high flow)

Cross-referencing (redundancy)

As part of a distributed metering system

Estimation Methods

Gas produced

Difference = Gas Flared

Exported gas

Fuel gas

Re-injected gas

More accurate during high flaring (meter back-up)

Unsuitable for low, background flaring…

Here’s a simple example to prove this....

By-difference Method

Assume, for simplicity, Uncertainty = 5% on both

gas in and gas out figures

Gas in = 100 t/hr 5% ( 5 t/hr)

Gas out = 60 t/hr 5% ( 3 t/hr)

Gas flared = 100 – 60 = 40 t/hr

% Unc = (52 + 32)0.5/40 x 100

= 14.6%

Example 1: High Flare

5%

5%

Gas in = 100 t/hr 5% ( 5 t/hr)

Gas out = 99 t/hr 5% ( 4.95 t/hr)

Gas flared = 100 – 99 = 1.0 t/hr

% Unc = (52 + 4.952)0.5/1.0 x 100

= 704%

or -6 to +8 t/hr

Example 2: Low Flare

Incorrect methods & equipment used

Poor maintenance & lack of traceability

Loss of signal at low/high flow (resolution/noise)

Bad installation and resulting errors/noise

Liquids, sand, scale… (especially at high flow)

Inadequate measurement uncertainty calculations

Poor spreadsheet analysis (incorrect filtering)

Inadequate measurement and use of p, T & Mw etc..

Inadequate data acquisition

General Measurement Issues

Some real-life Examples

Typical Meter Output

1) Background flaring

2) Routine Flaring

3) Upset condition

Area under graph = Total gas flared

Same data plotted cumulatively

Flo

w (

kg/h

r)

250

200

150

100

50

0

1 2 3 4 5 6

(days)

High Flow (meter over-ranged)

Wide variation with sudden spikes in data

Mass flow tops-out at 250 kg/hr (20 mA)

Low Flow Issues (refinery flare)

Negative velocity!!

Positive velocity

Courtesy: GE Sensing

Velocity on

various meter

paths

Temp & dew

Point plot

BY CALCULATION

PROCESS MODELLING

SOFTWARE

• „‟Advocated‟‟ in EU ETS MRR

• Requires knowledge of:

All inputs

Relative percentages

Process conditions

• Depends on reliable and

accurate process modelling

software!

• Models need to be “trained”

BY MEASUREMENT

GAS SAMPLING

• Manual sample (Flare-line):

H&S issues x

low pressure x

contributing streams

• Automated sampling Cost??

* baseline

* normal

* blow-down ?

contributing streams

ULTRASONIC

FLARE GAS METER

• Estimate gas composition

from speed of sound:

(average molecular weight)

* Use to estimate nitrogen

content (purge)

Flare Gas Composition

Mole-weight output (FG US meters) used in Norway to

enable corrections (reduction in tCO2) by nitrogen injection

Not widely used in North Sea, or worldwide

Could provide a method of determining Mw (and CO2

emission factor) during process upsets

Case Study – Using CFD Simulation to

Determine Installation Error in a Flare Gas

Flow Meter

Offshore operator was unhappy with the current

metering systems on two of their North Sea platforms

Company invested in a new ultrasonic metering system

on one of the platform to replace dual metering system

Meter welded to the flare boom pipework – there were

two main issues with this:

1. There was not enough straight length upstream to

ensure developed flow [carried out by NEL]

2. Meter required a heat shield to protect against

radiant heat from flare tip [addressed by the meter

manufacturer]

Case Study

Most flare meters are retrofitted to existing lines

Space/access is often a problem offshore

Shutdown to upgrade costs lots of ……

Installation Error

Result: meters installed in short pipe-runs

Bends, reducers, tees etc. cause:

flatness, asymmetry, swirl, and increased

turbulenceInstallation

error

Uncertainty

Installation Error

Flow

Meter

Good

Direction of flow

Error /

Increased

Uncertainty

Bad Flow

Meter

0

0.2

0.4

0.6

0.8

1

1.2

-1 -0.5 0 0.5 1r/R

vel

oci

ty

0

0.2

0.4

0.6

0.8

1

1.2

-1 -0.5 0 0.5 1

r/R

vel

ocit

y

0

0.2

0.4

0.6

0.8

1

1.2

-1 -0.5 0 0.5 1

r/R

vel

oci

ty

0

0.2

0.4

0.6

0.8

1

1.2

-1 -0.5 0 0.5 1

r/R

vel

ocit

y

Fully Developed Peaked Flattened Skewed

Asymmetric velocity profiles

Flow Profile and Swirl Patterns

Swirl patterns

Cannot install a flow conditioner owing to pressure drop

Change metering system? (Costly, need shutdown)

Determine a correction factor(s) to remove error:

• Bespoke lab tests? (Costly, representative?)

• In situ gas tracer tests? (Costly, flow stability)

• CFD simulation is cost-effective, and generally

more informative and enlightening than testing,

but a good knowledge of CFD modelling, flow

measurement is needed

• Validation is important

Remedial Actions

Proposed

Meter

Position

CFD Analysis – offshore flare (1)

3d mesh used to represent the

flow to through meter

Proposed

Meter

Position

Flow split

varied

Disturbed flow

profile + high

swirl = Error

CFD Analysis – offshore flare (2)

CFD (Flare Ultrasonic meter)

Single bend

Ideal, fully

developed flow

Disturbed flow

Dual-path research meter tested and

simulated at NEL (NMO funded project)

CFD Validation work

(Research Supported by NMO)

Flare is a uniquely difficult measurement application:

Very wide turn-down, erratic behaviour, low

temperatures, unstable flow

Hot-tapped meters cannot be removed for calibration

Installation errors, liquids, very low flow etc.

What uncertainty can we realistically expect to

achieve? (facility-specific: onshore vs offshore)

Practical, but comprehensive guidance is needed

Summary (1)

Summary (2)

Sampling from flare lines is difficult

Samples of fuel gas ok if mol weight during process

upsets can be characterised/estimated (taking into

account inorganic gases: N2, He etc.

Must take ratio of low/high flare into account (i.e.

contribution to total gas flared)

Direct measurement techniques (optical etc.) to

characterise emissions (costly)

Multiple meter technologies / methods may provide a

solution to complex installations

Examine flow and composition data properly – must

involve measurement experts to advise environmental

managers and consultants

Legislation, with cost implications (e.g. EU ETS), will

hopefully help drive innovation

Gas price dictates the value in capturing gas (e.g.

fracking in US has driven down the gas price

Summary (3)

Thank you for listening

Any questions?

The National Measurement System is the UK‟s national infrastructure of measurement

laboratories, which deliver world-class measurement science and technology through four

National Measurement Institutes (NMIs): LGC, NPL, the National Physical laboratory, NEL the

former National Engineering Laboratory, and the National Measurement Office (NMO).