Anders Quitzau, Innovation Executive, IBM

28. maj 2014

Transforming Energy and Utilities

through Big Data & Analytics

2014 IBM Corporation

What is changing in the Energy and Utilities industry?

Alternative energy enters the mix

The grid gets older

60% of electric grid
assets will need
replacement in this
decade

Smart meters become mainstream

In 2020, all Danish
electric utilities will
have smart metering
infrastructure (AMI)
installations

Global installed wind

power capacity
increased by 12.4
percent to more than
318 gigawatts in 2013

Consumers want more control &

insight into energy consumption
The number of U.S.
customers capable of
accessing information on
their energy use online
increased from 5.4M in
2010 to 17.5M in 2012

2014 IBM Corporation

To sustain growth, leaders across energy & utilities are prioritizing

four imperatives
Transform the
utility network

Continually improve
operational excellence

Improve generation
performance

Transform customer
operations

2014 IBM Corporation

Big Data is not just data, there are a few new considerations
Open Data

Big Data
Volume

Velocity

Variety

Veracity

Visibility

Data at
Rest

Data in
Motion

Data in
Many Forms

Data in
Doubt

Data in
the Open

Terabytes to
exabytes of
existing data to
process

Streaming data,
milliseconds to
seconds to
respond

Structured,
unstructured, text,
multimedia

Uncertainty due to
data inconsistency
& incompleteness,
ambiguities, latency,
deception, model
approximations

Open data is
generally open to
anyone. Which raises
issues of privacy.
Security and
provenance

Value

Data of
Many Values
Large range of data
values from free (data

philanthropy to high
value monetization)

Big data is defined by IBM as any data that

4

cannot be captured, managed and/or processed

using traditional data management
components and techniques
2014 IBM Corporation

Big data makes a big difference

Organizations using big data and analytics are up to

23x
more likely to report they are

substantially outperforming
their competitors
than those who do not use big data and analytics

Source: Analytics: The real-world use of big data, a collaborative research study by the IBM Institute for Business Value and the Sad Business
School at the University of Oxford. IBM 2012
2014 IBM Corporation

Using data, outperformers excel

in three key areas

108

108

more likely

more likely

Make data
accessible

more likely

57%

31%

26%

Draw insights
from data
Underperformers

84

54%

54%

26%

Outperformers

Translate
insight into
action

2014 IBM Corporation

Source: Leading through connections, Insights from the 2012 Global Chief Executive Officer study, IBM Institute for Business Value IBM 2012

Big Data & Analytics capabilities help utilities leverage data to meet the
challenges of a connected world
Data is emerging as
the worlds newest
resource for
competitive
advantage

Decision-making is
moving from the elite
few to the
empowered many

As the value of data

continues to grow
current systems
wont keep pace

Data explosion

Cost and pricing pressures

Increasing consumer
expectations and concerns

In the utilities industry, the

number of connected devices
participants in the Internet of
things is growing
exponentially:

By 2050, the Electric Power

Research Institute estimates
that the average electric bill
will probably go up by about
50 percent if the smart grid is
deployed.

45%
Compound annual growth
rate, 2010-2015

ca. 400%
is the expected increase in the
average electric bill if the smart
grid is not deployed

>50% of surveyed consumers

with an opinion expect smart
grid technologies will lower total
household costs for energy use.
But . . .

49%

of consumers were concerned

that erroneous smart meter
readings would result in
overcharges
2014 IBM Corporation

Big data & analytics are transforming energy and utilities

Energy and Utilities are turning knowledge into power by using

big data & analytics to better understand and shape customer
usage, improve service levels and availability, and detect and
prevent energy theft.

2014 IBM Corporation

Big data & analytics capabilities are required to address these

challenges and opportunities
New/Enhanced
Applications

All Data

SCADA

ERP/ MRP

Telematics

Customers

What is
happening?

Real-time Data Processing & Analytics

Operational
data zone

Landing,
Exploration
and Archive
data zone

Call
Centers

Discovery and
exploration

Deep
Analytics
data zone

What action
should I
take?
Decision
management

EDW and
data mart
zone

What did
I learn,
whats best?
Cognitive

What could
happen?
Predictive
analytics and
modeling

Social
Media

Why did it
happen?
Reporting and
analysis

Fraud
Mgmt

Warranty /
Quality

Outage
Management

Regulatory

Information Integration & Governance

Location

Systems

Security

Storage

On premise, Cloud, As a service

Grid

Asset
Management

2014 IBM Corporation

Key usage patterns for big data & analytics have emerged

Grid
Operations

10

Smart
Metering

Asset &
Workforce
Management

2014 IBM Corporation

Grid
Operations

11

2014 IBM Corporation

Big data & analytics capabilities can drive real business value from
Grid Operations
MV & LV Grid
Planning
System Control

Asset Management

Grid
Operations

Events & alarms

Fault location
Grid conditions
Load History
Temperature & Condition
Ratings & Impedance

Voltage Regulation &

Protection

Voltage & PF profiles

Batteries & secondary equip
Exceptions & Excursions

Customer Operations

Improved power quality

Fewer outages & events

Field Services

Regulatory Reporting
12

Detailed Load & voltage data

Embedded generation impacts
Load & Phase Balance
Topology & switching history

Avoid manual collection

Load & voltage surveys
PQ & outage investigations
SAIDI/SAIFI
Improved risk assessment response
Lower technical and non-technical
losses
2014 IBM Corporation

Enhance Grid Operations

All Data

IBM
Analyze data from continuous
sources in real time. Import and
execute predictive models.

Watson

Discover and navigate through

structured and unstructured
Foundations data across the entire
New/Enhanced
enterprise
Applications

SCADA

Increase generation capacity in

Real-time Data Processing & Analyticsresponse to demand

AMI

High bandwidth time series data

ingestion and recall
PMU

Operational
data zone

Landing,
Exploration
and Archive
Cleanse and validate data
data zone
from multiple customer or

Location

outage
EAM

management systems
prior to loading in the
warehouse

Social

ERP/ MRP

Demand
Forecasting

Discovery and
exploration

Deep
Analytics
data zone

What action
should I
take?
Decision
management

EDW and
data mart
zone

What did
I learn,
whats best?
Cognitive

What could
happen?

Deep analysis of energy

demand and loads and the
impact of distributed generation
Information Integration & Governance resources
Systems

What is
happening?

Monitor the performance and

status of the grid via reports
and dashboards.

Security

Predictive
analytics and
modeling

Storage

Why did it
happen?
Reporting and
analysis

Load
Forecasting

Distribution
Mgmt

Outage
Mgmt

Energy
Trading

Predict excessive loads on the

grid before they occur

On premise, Cloud, As a service

13

Master relationships and

build / maintain a single
view of the grid

Strong real-time, streaming

data performance across
distributed grid data

2014 IBM Corporation

Big data & analytics capabilities are transforming Grid

Operations
How can I uncover
anomalies in cell relay
signals to predict and
prevent power outages?

How can I manage the

load capacity of the
grid without relying on
manual calculations ?

How do I integrate
renewable energies on
the grid and forecast
energy consumption ?

Anticipate outages

Manage the flow of

power through the
grid automatically

Simulate energy
demand

Australian Power
Company

Utility company
in France

Energy utility
company in the
United States
Synthesized and analyzed
a large stream of data from
5,500 cell relays and 2.3
million smart meters
Predict and prevent
operational issues
Respond to outages more
efficiently by dispatching
crews to the right place at
the right time
14

Implemented a rules
engine and advanced
analytics to continually
calculate the theoretical
load limits of assets within
the grid network

Forecasts national
demand every 30 minutes
for a full year in advance,
versus daily in the past

Extended asset life and

subsequent deferred
unnecessary capital
investment via better asset
load ratings.

35 million load curves

analyzed and modeled in
near-real time

2014 IBM Corporation

Smart
Metering

15

2014 IBM Corporation

Big data and analytics capabilities drive real business value from

Smart Meter data

Grid Operations

Smart
Metering

16

Reduce Outages &

Downtime
Detect Energy Loss,
Theft & Fraud

Field Service

Optimize maintenance
& repair

Resource Planning

Improve load
forecasting & demand
accuracy

Customer Service /
Customer Operations

Improve acceptance
rates for targeted
offerings

Regulatory
Compliance

Efficient Regulatory
Compliance
2014 IBM Corporation

Big data & analytics turns Smart Meter data into actionable
insight
How can I give
customers more control
over their energy use ?

How can I manage the

load capacity of the
grid without relying on
manual calculations ?

How can I use smart

meter data to reduce
energy consumption?

Help customers
reduce energy
consumption

Manage the flow of

power through the
grid automatically

End-to-end
management of the
grid

US distributor &
transmitter of
electricity
Used smart meter data for
a customer web portal,
where customers can
access their personalized
information and learn how
much electricity
Reduced load times by
more than 95 percent
Decreased query times
more than 97 percent
Reduced TCO from 1.3 TB
to 350 GB
17

Large California
Utility
Gathered & analyzed smart Implemented a smart meter
infrastructure that provides
grid data representing
real-time, integrated view of
diverse terrain, weather &
the grid
demographics
Will reduce demand on the
50% drop in short-term peak grid by about 1,000
loads
megawatts
15% drop in overall peak
Will reduce greenhouse gas
loads
emissions by at least
10% reduction in electricity
365,000 metric tons per year
bills
2014 IBM Corporation

Gain insight from Smart Meter big data

IBM Watson

All Data

Discover and navigate

through structured and
Foundations
unstructured data across the
entire enterprise

Collect, store and analyze

Complex query processing of
up to 50,000 data points/
historical meter usage data.
Meters
secReal-time Data Processing & Analytics

ERP

Location

Customers

Grid

Operational
data zone

Landing,
Exploration
and Archive
data zone

Master relationships and

build / maintain a single
view of the grid

Deep
Analytics
data zone

Analyze customer
energy usage, detect
What is energy theft or meter
happening?
tampering
Discovery and
exploration

What action
should I
take?
Decision
management

EDW and
data mart
zone

What did
I learn,
whats best?
Cognitive

What could
happen?

Offer TOU pricing or

demand / response
offerings

Predictive
analytics and
modeling

Information Integration & Governance

Systems

Cleanse and validate data from

multiple customer, billing or outage
management
systems prior to
18
loading in the warehouse

Security

New/Enhanced
Applications

Storage

On premise, Cloud, As a service

High availability of data and

analytics processing of
customer preferences

Customer selfserve portals

Fraud / theft
protection

Why did it
happen?
Reporting and
analysis

Call Centers

Outage
Mgmt

Billing

Predict which customers

systems
are candidates for TOU
pricing or demand /
response offerings

2014 IBM Corporation

Asset &
Workforce
Management

19

2014 IBM Corporation

Asset & Workforce Management drives real business value for

utilities
Easy access to all documents
needed to operate the assets

Operate the Assets

Maintain accurate As-Built

documentation

Enforce business rules for

asset operations

Asset &
Workforce
Management

Populate work packages with

complete and accurate
information

Maintain the Assets

Use analytics to perform

Predictive Maintenance &
Condition Based Maintenance

Schedule Use of
the Assets
20

Capture changes so As-Built

configuration is accurate

Schedule maintenance based

on equipment conditions, not
fixed schedules
Minimize planned and
unplanned maintenance
activity
2014 IBM Corporation

Enhance Asset & Workforce Management with big data & analytics
Protocol
conversion, message mediation &
transformation of grid or power
All Datageneration data

Real-time analysis of
grid / generation asset
data

SCADA

Landing zone for data

collected from grid assets
DCS for off-line analytics

Grid

EAM

Operational
data zone

Pre-built data schema

for quality, machine and
production data

Social
Media

happening?

Deep
Analytics
data zone

What action
should I
take?
Decision
management

EDW and
data mart
zone

What did
I learn,
whats best?
Cognitive

Why did it
happen?
Reporting and
analysis

What could
happen?

Recommend preventative
maintenance

Systems

Load
Forecasting

Discovery and
exploration
Mobile
workforce

Asset Mgmt

Outage
Mgmt

Predictive
analytics and
modeling

Information Integration & Governance

Consolidate grid data for a

ERP/ MRP
360 degree view of the grid
or generation plant

PMQ
Capabilities
21

Landing,
Exploration
and Archive
data zone

New/Enhanced
Applications

Monitor the performance of

equipment and assets during
What
is
operation

Real-time Data Processing & Analytics

Internet scale messaging

Telematics

IBM Watson Foundations

Warranty /
Quality

Security

Storage

On premise, Cloud, As a service

Define and manage the product
master data

Predict an equipment failure and

avoid an outage

Strong real-time, streaming data performance

across distributed asset info

2014 IBM Corporation

Big data & analytics is transforming Asset & Workforce

Management
How do I reduce
unscheduled
maintenance costs?
Early failure
detection

Gas Turbine
Manufacturer
Used text mining methods
for extracting insights from
unstructured sources

Achieved 90% failure

prediction accuracy for gas
turbine compressor
subsystem
Identified 11 month early
warning of production and
supplier issues
22

How do I measure the

risk of failure and
optimize repair /
replacement work?

How can I minimize down

time related to asset
maintenance?

Asset & Workforce

Management
Optimization

Analytics Driven
Mining Asset
Management

Water Utility
Company

Large Mining
Company

Used advanced analytics to Performed analytics on data

from trucks / mining
identify potential problems
equipment, weather,
based on location, time,
operational performance and
weather and maintenance
ore price to assess asset
history
health
25%+ increased crew
utilization; 10-15% fuel
savings

Estimated $3B increased

profit on $30B operation

30-50% savings on
selected inspection and
preventive maintenance

Reduced 10-day failure

probability to less than 1%
2014 IBM Corporation

How to move strategically to transform your business

Build a culture that

infuses analytics
everywhere

Be proactive about
privacy, security
and governance

Enable and motivate your people

To trust the insights you have

to trust the facts. Big Data
also requires data
governance

Infuse analytics into key business

processes

Privacy and security to

protect the data

Deploy the full range

of analytics

Enable risk-aware decisions

Find the most compelling usecases and the business sponsor

23

Invest in a big data

& analytics platform
Build towards a platform for
all data and analytics
Analyze data in motion
Cultivate new partnerships
and roles

2014 IBM Corporation

ibmbigdatahub.com/industry/energy-utilities