AMI Lessons Learned

AMI Case Studies
and
Lessons Learned
Notes from the Field
Prepared by Tom Lawton,TESCO

Abstract
• As many utilities have elected to deploy advanced metering systems
and millions of new solid-state, microprocessor based end-points
with communications under glass, there has been a dramatic shift in
the approach to qualifying and certifying electricity meters.
• This presentation will highlight the need for a rigorous approach to
meter certification as well as continued testing as both the meter
and the communication module and the firmware to run both are
upgraded and changed over the course of deployment.
• Examples of issues which have arisen over the course of various
deployments along with varying approaches to and interpretations of
the same ANSI tests by manufacturers will be discussed.

AMI Deployment
Common Features and Common Sources of Concern
• Compressed time frames for pre-deployment, pilots,
and deployment, pilots, and deployment
• New features and functionality being added at a
tremendous rate
• Bugs, fixes, improvements being implemented
constantly in the meter
• Metering typically under represented or not
represented at all in AMI planning teams and
AMI deployment teams.
• The need for testing throughout this process is
rarely understood or even appreciated by AMI
planning and deployment teams.
• Issues around certification testing and acceptance
testing are initially seen simply as obstacles to
staying on schedule and not as valuable safeguards
against making potentially expensive mistakes

The Difference Between Meter Qualification, Meter
Certification & Meter Acceptance Testing
• Meter Qualification – the easiest and least time consuming aspect
of putting a new AMI technology in the field.
• Meter Certification – typically underestimated for the initial effort
and unanticipated over the course of deployment.
• Meter Acceptance Testing – more than just accuracy testing.
Includes functional, firmware and communication testing of all
incoming shipments.

Managing Change – at the utility, at the vendor
and from the suppliers to our vendor
• Every change is potentially important and should be treated this
way.
• There is no “small change”.
• There is rarely the same level of testing done for both the meter and
the communication device and never for the meter and the
communication module as an integrated device.
• Are ANSI tests without the communication module integrated always
suitable from a business, commission and customer view point.

And who is left standing that can do some
of these things at each step in the process?
Who Should be Doing What?
•Meter Vendors
•Communication device vendors
•AMI Team
•Metering Group
•Third Party test Labs
- Where are the risks?
- How closely should each
participant be held accountable
and at which steps?

Effective Methodologies
• Define the difference between each of the three types of testing and
certification
• Define the testing protocols to be used – talk to other utilities,
vendors, third party labs. Define your own and be prepared for them
to evolve over the course of your project
• Keep to the protocols.
Approvals, Failures
(do not term this “has
not passed yet”),
Deviations should all
be documented
DEFINEDEFINE
EVOLVEEVOLVE
FOLLOWFOLLOW

Typical Protocols:
• AMI Teams typically do a good job of defining what they expect the meters and
communications devices to do.
• Meter Services can take this specifications document and check to see if the
devices proposed can handle these tasks.
Ways to stay out of trouble and to set yourself up for success in the subsequent
steps:
• Define the testing protocols to be used – talk to other utilities, vendors, third
party labs. Define your own and be prepared for them to evolve over the
course of your project.
• Keep to the protocols. Approvals, Failures (do not term this “has not passed
yet”), Deviations should all be documented.
• Check for yourself. If a feature is advertised then you should be able to verify
the feature is functional and whether this feature meets your needs. When
purchasing a “Smart” meter, know what you are getting.
Qualification Testing

• ANSI Testing
• Meter Functionality
• Meter Data Acquisition Accuracy for Each Form
• Dual Socket Meter Comparison – Data Collection & Accuracy
(Energy, Demand, Load Profile)
• Large Test Platform Meter Comparison –
Data Collection & Accuracy
• Disconnect/Reconnect Functionality
• Outage Performance
• Meter Communications Performance
Pre-deployment Certification Testing

• ANSI Testing
• Meter Functionality
• Meter Data Acquisition Accuracy for Each Form
• Dual Socket Meter Comparison – Data Collection & Accuracy
(Energy, Demand, Load Profile)
• Large Test Platform Meter Comparison –
Data Collection & Accuracy
Deployment and Post-Deployment
Certification Testing

• Accuracy testing
• Meter Communications
Performance
• Software and firmware
verification
• Setting verification
• Functional testing
• Disconnect/reconnect
Functionality
and as left setting
Acceptance Testing

Questions that need to be asked:
• How does Acceptance Testing
differ from Certification Testing?
• When do we return to
Certification Testing?
• What should be built in to the
AMI Vendor Contracts?
• Who should be involved in
developing the Certification and
Acceptance Test Plans and
when should this be done?
Setting up Certification and
Acceptance Testing Plans

Useful Tools & Potential Resources
• ISO 9001
• Quality System Documentation if not ISO 9001
• Quality systems and test reports from vendors
• ANSI Test results, and protocols used by various
manufacturers for these tests. Insist upon
additional testing where you feel necessary
• Third Party Labs for your own certification testing

ANSI & Other Standards
• Not all Standards are created equal. The same Standard is
often not interpreted the same way by all test facilities. The
devil is always in the details. Do not just ask if the meter
passed a particular test, but learn the details about how this
test was run. Also find out,
• What version of the meter?
• Was the communication module
integrated with the meter for any
of these tests?

Vendors, Consultants, Independent Labs,
Other Utilities, Networking
• You are not the first Utility to go through this process. At this point several others
utilities have completed or are nearing completion of a deployment with your
meter of choice or one of the meters you are considering choosing. There may
even be utilities who have deployed the same meter with the same
communication device. Ask your peers. Pick their brains. Talk to the metering
folks. We are pretty friendly individuals.
• Most of the vendors have gotten better about no
longer stone walling with “we have never seen
that issue before”. We are asking our meter
vendors to deploy new technology and new
features very rapidly. They generally want to
learn of any issues as quickly as possible and
address these issues as quickly as possible.
• You always have a first or a final resort to go to
an independent test facility – new feature never
used in the industry or a problem that crops
up after deployment has begun

Quality Audits
• Schedule these for your meter manufacturer and your communication device
manufacturer.
• Objective is to understand how they tested the product in the first place as well as
how any concerns that come up will be addressed. What is the mechanism, who are
the people, what capabilities do they have to investigate and determine root cause.
How and when are these responses fed back to you the customer and to the
manufacturing and supply chain.
• In plant final inspection and testing
• Plant first pass test results… if the number is less than 95% (which is typical),
validating and auditing the off-line repair process becomes equally as important as
auditing the main line quality checks
• Vendor supply change management
• Validation testing of firm ware changes

RMA’s and Self-Audits
• In conjunction with auditing your vendor you should
audit your own procedures for identifying, reporting and
following up on issues.
• For many this is simply a review of their RMA
procedures. These procedures need to be
strengthened and personnel assigned to them as the
number of issues will skyrocket over the course of a
deployment. Most will be of no or little consequence.
Some will be important and one or more may have
catastrophic results if not addressed quickly and
followed-up forcefully and with persistence.
• Regardless of the extent of the self-audit, your quality
system will have to be strengthened and staffed to
handle the demands and the rigors of an AMI
deployment. Clear chains of command and reporting
protocols and decision matrixes will be as important as
any testing protocols.

Post-Deployment
Meter Farms as a Tool
• Once deployment is complete the certification and acceptance testing does not
stop.
• Everyone understands the importance of acceptance testing, but future
generations of a smart meter also require certification testing.
• One new method for managing existing meter populations as well as future state
meters is through meter farms.
• New software and firmware can be tested and compared against the performance
of older generations.
• New hardware can be tested and compared to older generations.
• Firmware upgrades can be checked
• Long term issues can be identified or verified in the
meter farm
• New head end systems or IT protocols can be
tested against the meter farm before going live
as well.

Real Life Examples of Change
Management in Action
• Issues in the system not the components – what is the definition of a meter
and when did the meter change?
– Meter and communication device interface
• Feature Creep – every manufacturer wants to differentiate themselves –
sometimes this works in unanticipated ways
– Recovery from power outages
– Short and long demand periods
• But we only changed….lessons we should
have learned from Microsoft
– Over the air upgrades
• Thank Goodness for testing plans and
contracts – right?
– Half closed disconnect devices
– Disconnect devices of unknown state
– Meters with incorrect firmware

Expected & Unexpected Trends Toward
the End of Deployment
Expected:
• The AMI deployment team will declare victory at some point and move on.
Clean-up will be left for the meter service department
• Change Management continues to be a tremendous challenge for every
vendor through every deployment
• Meter Acceptance testing including far more than accuracy testing for every
deployment
• Firmware upgrades must be checked and tested before mass deployment
Unexpected:
• Not all forms are available by the end of
deployment and must be installed as
part of normal operations by the meter
service department
• Meter Certification Testing never slowed
down over the course of any of the
deployments

Summary
• There are no bad manufacturers. Everyone is moving very
fast to meet the market demands. Without adequate checks
and balances there will be problems. Even with them, there
may still be problems – that is why we call this Risk Mitigation
and not Risk Elimination
• Each Utility must take a far more active role as part of this
system of checks and balances

10/02/2012 Slide 22
Customer Perceptions of New
AMI Meters & What
Meter Services Can do
to Help or Hinder the Cause
Prepared by Tom Lawton, TESCO
The Eastern Specialty Company

• Understand Customer concerns about
Smart Meters and AMI
• Be able to respond as a Utility and as a
Utility Worker
Session Objectives

• Health - new AMI meter’s will kill my cat or
make me sick
• Safety – new AMI meter’s will burn down
my house
• High Billing – new AMI meter is going to
increase my bill
• Privacy - Big Brother is watching me
Customer Concerns – Real & Perceived

• RF Emissions
• We sites and internet information and
misinformation
• Cell Phones vs Meters
• Measuring RF
Health Concerns

For the first time in our collective careers meters have been in
the popular meter in an unflaterring way. Segments of the
general population have the perception that;
• AMI Meters may spontaneously catch
• AMI Meters may blow up
• AMI Meters may disconnect power
by themselves
• AMI meters are “cheap computers” and are not
robust enough for long term outdoor use
Safety Issues?

When US Consumers think about electrical safety within their
home they think about licensed electricians and inspectors for
electrical work and the UL Mark for products used within their
home.
• Customer: Why don’t meters have a UL label?
Are the AMI meters safe?
• Utility: Are ANSI standards “tough” enough?
How good is the meter vendor testing?
- Excerpt from PECO News Release October 9, 2012
What Regulatory Alternatives Are Out There?

Currently a Draft Standard under review
Scope
All Type S and Type A electric meters rated up to 600
volts and any other type of meter intended for
installation within the enclosure of “complete
equipment”.
Contents
Meter Construction Requirements
Meter Performance
Meter Markings
Standards for Components
UL Meter Safety Standard 2735

Meter Construction Requirements
Meter forms as defined in ANSI C12.10 unless
alternate forms are specified by a Utility
Enclosures
Covers
CT’s, internal and external
Batteries
Service Switches
Circuit Boards
UL recognized or tested components

Meter Performance Tests
Tests for various fault conditions
How easy to set on fire
Strength of Construction
Some tests from ANSI C12.1 Section 4

Component Standards
UL Standards for many meter components will
apply:
• Fuses
• Transformers
• Switches
• Terminal Blocks
• Connectors

• UL is very interested in capitalizing on this opportunity to
regulate electric meters – a market they have been excluded
from in the past by the NFPA
• Draft Standard 2735 was issued without any prior notification to
the ANSI C12 Main Committee despite assurances that UL
would work with and participate with ANSI to avoid a Standard
that contradicts the complementary ANSI standard
• Manufacturing cost for a meter would increase significantly –
neither the manufacturer nor the end user are sure how much
this cost increase will be or if the increase is warranted.
Issues to Address

• Agreement on a common standard by UL and ANSI
• Meter vendor acceptance of the new standard
• Coordination of UL and ANSI testing of meters
• Lead Time and Cost of UL listing
• Need for ANSI Testing in the shadow of a UL Standard
• UL part of new meter certification process
• UL part of new component selection
and design changes
Challenges of Implementing a UL Standard

Pro’s
• Greater perception of safety by the general public
• Outside inspection to maintain certification
• All changes to meter construction are monitored
and approved by an outside group
Pro’s & Con’s of a UL Mark on Electric Meters

Con’s
• Greater cost to Utility and Utility customers
• Slower innovation for meters
• Potential for short term meter shortages after
implementation
• Potential for fewer meter vendors and options for
Utilities
Pro’s & Con’s of a UL Mark on Electric Meters

Current involvement of UL in the metering space
• Independent Test Lab to run ANSI tests
• Independent Test Lab to run customer specific tests
• Independent Test Lab to recommend and run safety
tests on any metering product
Use of UL as an Independent Test Lab

10/02/2012 Slide 37
Hot Socket Issues
Causes and Best Practices
Prepared by Tom Lawton
TESCO – The Eastern Specialty Company
with special thanks to L+G who sponsored much
of the research presented here today

The Issue
• Hot Sockets are not a new phenomenon. Virtually every meter man has pulled a meter with a
portion of the meter base around a blade melted and virtually every utility has been called to
assist in the investigation of a fire at a meter box.
• AMI deployments because of the volume of meters involved put a spot light on this issue.
– What causes a hot socket?
– Are the meters ever the cause of a meter box failure?
– What are the things to look for when inspecting an
existing meter installation?
– What are the best practices for handling potential hot
sockets?
• This presentation will cover the results of our lab
investigation into the sources for hot sockets, the
development of a fixture to simulate hot sockets,
the tests and data gleaned from hot sockets, and
a discussion of “best practices” regarding hot sockets.

Why do we know anything about hot sockets?
• TESCO has been fortunate enough to be involved in several meter deployments where we
supplied full time and part time meter engineers and project managers to our customer’s AMI
deployment teams. In this capacity we have been involved in evaluating hot socket issues and
helping to determine an appropriate response to actual or potential hot sockets.
• TESCO’s meter lab has been contracted to develop a laboratory fixture that would simulate the
various features common to most hot sockets found in the field.
• TESCO developed and refined a fixture since the 2013 Fall EEI running tests and gathering data
on the effect of hot sockets on meters.
• TESCO has access to a large number of
meters which have been exposed to hot
sockets both before and after catastrophic
failure.
• We have access to a limited number of
sockets that were hot sockets and did not yet
fail catastrophically.

Searching for Hot Socket sources
• Pitted and discolored meter blades
• Melted plastic around one or more of the
meter stabs (typically the plastic around
one stab is where the deformation starts)
• Pitted and discolored socket jaws
• Loss of spring tension in the socket jaws

Hot Socket Simulation Fixture
Click to view video

Expected & Unexpected Results
Expected:
• Hot Sockets are exactly that – hot sockets. The hot sockets are the source of the
problem and not hot meters.
• Electromechanical meters withstand hot sockets better than solid state meters
Unexpected:
• Current plays only a small role in how
quickly a meter will burn up. Meters were
burned up nearly as quickly at 3 amps, 30
amps, and 130 amps.
• Relatively small amounts of vibration can
be the catalyst in the beginning and
eventual catastrophic failure of a hot
socket. Note: Other catalysts include but
are not limited to power surges, debris,
humidity, salt water.
• Contact resistance plays no role in
creating a hot socket
And some newer solid state meters are
better than electromechanical meters.

Temperature Rise Data

Jaws with intermittent
connections will arc to the
meter blade resulting in
pitting on the blade.
Blade shows early signs
of arcing.
Tin Melts at 232ºC which
is lower than the 350ºC
base plate plastic.
Slide 45
Jaw to Blade Arcing

46
Tin burned off
Blade hole due to arcing
to jaw – Copper melts at
1040ºC (1900ºF)
AX-SD base thermoset
plastic melts at 960ºC
(1760ºF)
Severe Arcing Jaw to Blade

What are the necessary ingredients
for a hot socket?
There are three necessary ingredients to create a hot socket (Note: We are
not suggesting that we have simulated or even understand all causes for all
hot sockets and meter related fires, but rather that we have simulated and
understand the causes behind most hot sockets and meter related fires);
• Loss of jaw tension in at least
one of the socket jaws.
• Vibration (or other catalyst
to initiate arcing)
• Minimal load present

Reviewing the data and learning from the data
• Repeated meter insertions degrades the tension in the socket jaws (see graph), but
not to dangerous levels
• Exposure to elevated temperatures rapidly degrades the socket jaw tension to
dangerous levels (see graph)
• Visual inspection will catch some but not all dangerous socket jaws
• Arcing creates the heat
• Exposure to elevated temperatures has a cumulative effect on the meter socket jaw
• Relatively small vibration can initiate arcing

Field Inspection of Sockets
Best Practices
• Example field check list
– Gaps in meter socket jaws
– Discoloration of one jaw vs. the other three
– Signs of melted or deformed plastic on
meter base
– Pitting of either meter blade or socket jaw
– Loss of tension in meter socket jaws
– Check condition of wire insulation and
connections to meter jaws
– Check the overall condition of the box,
socket, meter and how they attach to each
other and the building.
– Look for signs of tampering
– Look for signs of water or debris inside of
the meter can

• Most AMI deployments utilize third party contractors to handle residential and
some self contained non-2S services.
• After to or prior to AMI deployments, Utility personnel typically see these
sockets
• Transformer rated meters typically handled by the meter service department
of the utility.
• Hot socket concerns with lever by-pass sockets used on 3-phase meters are
extremely rare.
Who Sees Hot Sockets?

• Easiest resolution is to replace the damaged jaw.
• Never try and repair a damaged jaw. The tension in the
damaged jaw will not return simply by taking a pair of
pliers and closing the jaw tighter.
• Either the entire box should be replaced or the damaged
jaw (assuming the wiring and other jaws are deemed
safe through the rest of the inspection.)
What can be done once a hot
socket is identified?

Base Line Data Electro Mechanical meters vs solid state vs the latest
generation of meters designed with hot sockets in mind
• At the start of our laboratory investigation the oldest electro mechanical meters
withstood hot sockets the best
• The latest vintage solid state meters withstood hot sockets the least.
• Over the course of the past twelve months some meter manufacturers have begun to
release 2S meters designed to withstand hot sockets and some have even begun to
put temperature sensing closer to the meter blades instead of only on the metrology
boards.
• One meter vendor’s service switch meter has used high temperature base plate
plastic since it was launched in 2008.)

• Calipers show a .01” gap,
with that size gap between
jaws and stabs we were
able to heat meter stabs
over 1000 degrees
Fahrenheit in a few
minutes.
• The rough spots you see
on the post-test jaw next to
the calipers are over .005”
high. This surface
degradation appears on the
stab as well.
• Between the two surfaces
you can have large gaps,
along with insulating by-
product of the arcing, that
can sustain heavy arcing in
a solid state.
Service Degradation
Slide 54

• In a representative side view of a .1” thick standard meter stab, you
can see how small these distortions appear relative to the thickness
of the stab, while creating an air gap large enough for significant
arcing.
Service Degradation
Slide 55

Summary
• Hot sockets start with a loss of tension in at least one of the meter socket jaws. This
loss of tension can be from a variety of sources that start as early as improper
installation or even “tight sockets”.
• Loss of tension is necessary to create the initial micro-arcing conditions.
• Sockets with repeated meter exchanges observed to have higher incidence of hot
socket issues and “booting” a meter may spring jaws even more.
• Vibration appears to be the most common catalyst to the micro-arcing that creates the
initial heat in a “hot socket”.
• The meter must have some power, but current is not a significant factor in how quickly
or dramatically a hot socket occurs
• The effects of vibration and weakened jaw are cumulative
• Meter Manufacturers have all been working on the design of their meters to better
withstand a hot socket. These new meters have better baseline performance than
even the older electro mechanical meters, but a hot socket will eventually burn up
even the most robust meter.

10/02/2012 Slide 57
– Site Verification –
Protocols & Tools for
Complex Metering
Installations

Premise
• Over much of the 20th century, utilities, regulators and customers each
relied upon lab and field meter testing efforts which were primarily focused
upon the accuracy of the watt-hour meter and demand register.
• This focus is now changing with overwhelming deployment of electronic
meters and significant deployment of AMR and AMI meters throughout the
installed base in North America.
• The focus has now shifted to the metering installation as a whole and not
the accuracy of the meter.

Field Testing
Electro Mechanical meters were
subject to registration errors caused
by mechanical issues with moving
parts resulting in either the loss of
revenue to the utility or over billing
for the customer. Some of the more
common problems were:
• Friction wear
• Gear mesh misalignment
• Retarding magnet failure
• Timing motors

Electronic Meters – new failure modes require new testing
and inspection methods
Electronic meters fail as do electro mechanical meters but differently
• Their overall life expectancy is not nearly the same
• Failure modes include drift (unexpected)
• Failure modes include catastrophic (expected)
• Failure modes include non-catastrophic but significant measurement error
modes
• Failure modes can include non-measurement issues which render the meter
ineffective or inaccurate for billing purposes.

Best Practices
• Residential vs Commercial
• Self-Contained vs Transformer Rated
• Follow the money and be as proactive as possible

• Double check the meter number, the location the test result and the meter record
• Perform a visual safety inspection of the site. This includes utility and customer equipment.
Things to look for include intact down ground on pole, properly attached enclosure,
unwanted voltage on enclosure, proper trimming and site tidiness (absence of discarded
seals, etc.)
• Visually inspect for energy diversions (intentional and not). This includes broken or missing
wires, jumpers, open test switch, unconnected wires and foreign objects on meters or
other metering equipment. Broken or missing wires can seriously cause the under
measurement of energy. A simple broken wire on a CT or VT can cause the loss of 1/3 to
1/2 of the registration on either 3 element or 2 element metering, respectively.
• Visually check lightning arrestors and transformers for damage or leaks.
• Check for proper grounding and bonding of metering equipment. Poor grounding and
bonding practices may result in inaccurate measurements that go undetected for long
periods of time. Implementing a single point ground policy and practice can reduce or
eliminate this type of issue.
• Burden test CTs and voltage check PTs.
Potential Site Check List

• Verify service voltage. Stuck regulator or seasonal capacitor can impact service voltage.
• Verify condition of metering control wire. This includes looking for cracks in insulation, broken wires,
loose connections, etc.
• Compare the test switch wiring with the wiring at the CTs and VTs. Verify CTs and VTs not cross
wired. Be sure CTs are grounded in one location (test switch) only.
• Check for bad test switch by examining voltage at the top and bottom of the switch. Also verify
amps using amp probe on both sides of the test switch. Verify neutral connection to cabinet
(voltage).
• Check rotation by closing in one phase at a time at the test switch and observing the phase meter
for forward rotation. If forward rotation is not observed measurements may be significantly impacted
as the phases are most likely cancelling each other out.
• Test meter for accuracy. Verify demand if applicable with observed load. If meter is performing
compensation (line and/or transformer losses) the compensation should be verified either through
direct testing at the site or by examining recorded pulse data.
Potential Site Check List (cont)
• Loss compensation is generally a very small percentage of
the overall measurement and would not be caught under
utilities normal high/low checks. However, the small
percentages when applied to large loads or generation can
really add up overtime. Billing adjustments can easily be in the
$million range if not caught early.

• Verify metering vectors. Traditionally this has been done using instruments such as a circuit
analyzer. Many solid state meters today can provide vector diagrams along with volt/amp/pf and
values using meter manufacturer software or meter displays. Many of these desired values are
programmed into the meters Alternate/Utility display. Examining these values can provide much
information about the metering integrity. It may also assist in determining if unbalanced loads are
present and if CTs are sized properly. The vendor software generally has the ability to capture
both diagnostic and vector information electronically. These electronic records should be kept in
the meter shop for future comparisons.
• If metering is providing pulses/EOI pulse to customers, SCADA systems or other meters for
totalization they also should be verified vs. the known load on the meter.
• Verify meter information including meter multiplier (rework it), serial number, dials/decimals, Mp,
Ke, Primary Kh, Kr and Rate. Errors in this type of information can also cause a adverse impact
on measured/reported values.
• Verify CT shunts are all opened.
Potential Site Check List (cont)

• Accuracy Testing
Performance
• Software & Firmware
Verification
• Setting Verification
• Functional Testing
• Disconnect/Reconnect
Functionality
and as left setting
Shop Testing

Testing Frequency & Cost
• Tools (hardware and software)
• Personnel
• Frequency
• Test
• Report
• Analyze
• Learn, Share, Adapt

10/02/2012 Slide 67
Meter Operations
in a Post AMI World
for the Long Island Power Authority (LIPA)

Abstract
• As many utilities have elected to deploy advanced metering systems
and millions of new solid-state, microprocessor based end-points
with communications under glass, a dramatic shift has begun
regarding where metering resources are being deployed and what
they are doing.
• This presentation will highlight the new value proposition for
metering personnel at their respective utility companies in a post-
AMI World.
• Examples of issues which have arisen or been identified over the
course of various deployments and in the immediate aftermath of an
AMI deployment.

Pre-AMI Meter Operations
• Fewer meter techs in the field and in the shop than
there were 25 years ago
• Fewer Field checks and site verifications due to lack of
personnel, lack of experience and lack of expertise
• More features under glass in the meters even before
AMI deployed
• Significantly more features under glass in every AMR
and AMI system being considered or being deployed
• Metering losses starting to be identified “by accident” as
opposed to being “by design”.

The Pendulum Starts to Move
• Meter Operations. Prior to deployment many larger utilities take an attitude
of “this is only metering – how hard can that be?”. Over the course of
deployment Meter Operations gains a stature and a respect from the rest of
the organization that was not previously accorded them – even if this is only
begrudging respect.
• New tests for AMI meters (e.g. communication,
disconnect) are identified
• The complexity and issues around high revenue
metering are at least acknowledged by the rest of
the organization
• Tests which have not been performed in years are
suddenly back in vogue

The New Realities
• Electro-Mechanical Meters typically lasted 30 years and more. Electronic AMI
meters are typically envisioned to have a life span of fifteen years and given the
pace of technology advances in metering are not expected to last much longer
than this.
• This means entire systems are envisioned to be exchanged every fifteen years
or so. In the interim the meter population and communication network inherent
in the infrastructure for each utility must be maintained.
• Meter communication and meter data
management are becoming as important to
metering operations as meter accuracy.
• Firmware upgrades, firmware stability and cyber
security are becoming increasingly important to
metering departments

Expected & Unexpected Trends at the
End of Deployment
Expected:
• The AMI deployment team will declare victory at some point and move on.
Clean-up will be left for the meter service department
• Change Management continues to be a tremendous challenge for every vendor
through every deployment
• Meter Acceptance testing including far more than accuracy testing for every
deployment
• Firmware upgrades must be checked and tested before mass deployment
Unexpected:
• Not all forms are available by the end
of deployment and must be installed
as part of normal operations by the
meter service department
• Meter Certification Testing never
slowed down over the course of any of
the deployments

Why do AMI meters fail?
Looking back at various deployments – what are the chief
causes to reject meter shipments?
Meter functional test failures including but not limited to;
• Incorrect firmware
• Bad settings
• Alarms and errors that do not clear
• Communication test failures
• Bad tables
• Failed disconnect switches

Lessons for the Future?
• Are these infant mortality issues or are these issues to be concerned about
going forward?
• What was NEVER the reason for rejecting a shipment – meter accuracy.
• Which are the most difficult meters for every deployment – the transformer
rated meters. Often these forms are not available until very late in the
deployment and sometimes not until the deployment has officially ended.

• Most AMI deployments utilize third party contractors to handle the
residential and some self contained non-2S services. The balance are
typically handled by the meter service department of the utility.
• No AMI deployment has used the AMI communication network to handle the
communication with the largest customer meters. The risk from even a
short term interruption of communication or loss of data far outweighs to
benefit of meters which are already being communicated with daily or even
several times a day.
• As these services are evaluated for new metering technology issues are
being found at some accounts. These issues represent revenue losses due
to inappropriate metering schemes or partially failed metering components
(e.g. transformers, electronic components).
Commercial and Industrial Metering Challenges

• Metering is becoming more about IT. Some metering departments have
been reorganized after AMI as part of the IT department.
• Metering emphasis will shift strongly toward C&I customers and further and
further away from residential meters
• Metering resources are being refocused on C&I accounts.
• There are fewer levels of meter tech. Every meter tech needs to be at the
higher level as there is not enough lower level work to warrant full time
employees
• Utility commissions are being less forgiving of allowing rate relief for project
over runs and metering inaccuracies
How are these new realities starting to affect
metering operations?

• Responsible for either reviewing ANSI Tests or even performing some of
these ANSI Tests
• Perform Meter Functionality testing on new and returned AMI meters
• Register and communication module energy measurement comparison
• Consumer safety and combating real and perceived issues
• Near continuous research into the “next” technology and the next
deployment
New or expanded roles for metering
departments of all sizes

• Accuracy testing
Performance
• Software and firmware
verification
• Setting verification
• Functional testing
• Disconnect/reconnect
Functionality
and as left setting
Shop Testing

Field Testing
• Accuracy Testing
• Software & Firmware Verification
• Setting Verification
• Functional Testing
and as left setting
• Tamper Verification
• Site Audits appropriate to the type
of meter

RMA’s, Root Cause Analysis and
In Service Testing
Opportunities to Learn:
• Meters returned from the field as bad or flagged on incoming inspection. These
RMA’s need to be tracked not only for warranty purposes but more importantly to
understand the root cause of the issues.
• In Service Test programs need to be utilized to understand
the overall performance of AMI meters and as failures are
identified these meters need to be dissected and the
failure mode understood. If necessary, once the failure
mode is understood the meter population may need to also
be dissected to identify sub-groups of meter that may be
similarly affected. This could be a group of meters…
• from the same shipment.
• that were deployed in the same geographic area.
• that saw the same type of usage or environment.
No one knows the actual life of their AMI meter. To do this we must learn as
much as we can from the failures and the performance of the meters we have
deployed.

Post-Deployment Needs
• Once deployment is complete the certification and acceptance testing does not stop.
• Everyone understands the importance of acceptance testing, but future generations
of a smart meter also require certification testing.
• New software and firmware needs to be tested and compared against the
performance of older generations.
• New hardware must be tested and compared to older generations.
• Firmware upgrades need to be checked
• New head end systems or IT protocols need to be tested against a large
group of meters before going live.

Real Life Examples of Change
Management in Action
• Issues in the system not the components – what is the definition of a meter and
when did the meter change?
– Meter and communication device interface
• Feature Creep – every manufacturer wants to differentiate themselves –
sometimes this works in unanticipated ways
– Recovery from power outages
– Short and long demand periods
• But we only changed….lessons we
should have learned from Microsoft
– Over the air upgrades
• Thank Goodness for test plans – right?
– Half closed disconnect devices
– Disconnect devices of unknown state
– Meters with incorrect firmware

Summary
• The focus is on far more than accuracy
• The meter man of the future knows not only metering but they are communication
experts and are involved with handling, reporting and analyzing far more customer
and system data
• There are fewer “routine jobs” in the field and many of the standard tasks are no
longer required. This will lead to fewer classifications of meter techs going forward
and the need for a more highly trained tech.
• Each Utility must take a far more active role as part of checking, certifying, and
rechecking the functionality of their meters.
• Metering will be in a near constant cycle of looking for
the next technology, evaluating those technologies,
planning for deploying these technologies, and
cleaning up the aftermath of the deployment of
these technologies.

Questions and Discussion
Tom Lawton
TESCO – The Eastern Specialty Company
Bristol, PA
1-800-762-8211
This presentation can also be found under Meter
Conferences and Schools on the TESCO web site:
www.tesco-advent.com

AMI Lessons Learned

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (14)

Similar to AMI Lessons Learned

Similar to AMI Lessons Learned (20)

Recently uploaded

Recently uploaded (20)

AMI Lessons Learned