Industrial IOT

(Industrie 4.0)
By : Wan Razaq

Monday 27-08-2018
Outline
PREDICTIVE MAINTENANCE 4.0

WIRELESS STEAM TRAP MONITORING

 Table of Content
1. PREDICTIVE MAINTENANCE 4.0
1.1 Industry Landscape – Asset Intensive Industry
1.2 Type of Assets
1.3 Painpoints – Downtime in the asset-intensive industry
1.4 Maintenance Approach
1.4.1 Global Survey – Maintenance Programs adoption
1.4.2 Global Survey – Industry using predictive maintenance
1.4.3 Global Survey – Industry using analytics
1.4.4 Global Survey – Total PdM Market
1.4.5 Global Survey – Predictive Maintenance maturity level
1.5 Current maintenance technique - Condition Monitoring
1.5.1 Vibration Monitoring
1.5.2 Infrared Thermography
1.5.3 Oil Analysis
1.5.4 Acoustic Emission
 Table of Content
2. WIRELESS STEAM TRAP MONITORING SYSTEM
2.1 Industry Landscape – The use of steam in Industrial Facilites
2.2 Condensate
2.3 Steam Trap
2.3.1 Steam Trap Location – Tank Farms
2.3.2 Steam Trap Location – Flare Stacks
2.3.3 Steam Trap Location – Steam Turbines
2.4 Painpoints – The need to monitor Steam Trap
2.5 Current monitoring Technique
2.6 Solution Example - Emerson
PREDICTIVE MAINTENANCE 4.0 (PDM)
 1.1 Industry Landscape - Asset-Intensive Industry
requires large amounts of investment to produce a good/service, and therefore
have high percentage of fixed asset, including property, plant and equipment.

Painpoints and Challenges

1. The needs for larger capital investment and expenditure
Mining Chemical Power Generation Cement for deploying and maintaining asset.

2. The need to service and maintain assets that are often

geographycally distributed over wide areas as well as remote
and hard to reach.

Upstream O&G Refinery Fertilizer Textile

3. The need to operate in environments that are hars, volatile
and exposed to the elements.

4. The need to maintain these assets continually operating in

an optimal manner.

Automotive Pulp & Paper Aerospace Shipping Train

 Big names in Asset Intensive Industry

Mining Chemical

Cement
Power Generation
 Big names in Asset Intensive Industry (2)

Upstream O&G Refinery

Fertilizer Textile
 Big names in Asset Intensive Industry (3)

Automotive

Aerospace Shipping Train

1.2 Type of Assets
Asset in industrial can be divided into 2 group, rotating and stationary
equipment.
Rotating Equipment Static Equipment
A device that
Device that moves regulates, directs or
fluids (liquids or gases)
Valve
controls the flow of a
Pump or sometimes slurries fluid
by mechanical action
A closed vessel in
which fluid (generally
Boilers
A mechanical device water) is heated
that increases the
Compressor pressure of a gas by
reducing its volume A device used for high-
Furnace temperature heating

A rotary mechanical
A container designed to hold
device that extracts Pressure
Turbine gases or liquids at a pressure
energy from a fluid
Vessel substantially different from
flow.
the ambient temp.
 Open Pit Mining

Escavator &
dump truck

Source :Geo Energy Group

Oil & Gas

Compressor

Separator

Source : Corken
Oil & Gas distribution chain

Oil / gas
pipeline
Power Generation
Gas Turbine

Condenser
 1.3 Painpoints - Downtime in the asset-intensive industry
Downtime influences factors like equipment, labor overhead, maintenance,
engineering and production.
Tangible Costs Intangible Costs
Lost Production Lost Capacity Responsiveness Stress
• Every product that • When there is sudden • The cost of downtime in • Downtime can cause a lot
manufacturer produces increase in demand, the manufacturing is so of stress in both employees
represents some amount of factory will need to operate significant. and the machines that they
potential profit. at a higher capacity. are operating.
• It becomes more important
• Reducing downtime is to solve these problems
important because it than focusing on customer
creates additional capacity. service issues.

Direct Labor Inventory Innovation

• When you reduce • One cause of downtime is • Downtime can be a very
downtime, your production changeover. time-consuming.
levels go up while labor
• Reduce changover will • This takes away time from
stays the same.
allow smaller lot sizez and other things like inovation
• Hence, downtime will lower inventory levels. and creative brainstorming.
increase labor cost per unit.
 True Downtime Cost (TDC)
Includes a vast array of wasted business cost and lost business opportunity
because resources were needed to rectifity a downtime incident that shouldn’t
happen. True Downtime Cost
800

Downtime in Automotive Industry 700

In 2006, automotive industry executives showed 600

stopped production costs avg 22.000 USD per 500
minute. (1.3M per hour of downtime)

Hour
400

300

200
The Solution
100
53% of executives all agreed – reduction of
0
downtime can be done through predictive and
Automotive Food & Injection Machinery Metal Other Paper
preventive maintenance Bevarage Molding / and Products
Tobacco Extrusion equipment

Average Total Annual Downtime for industry Average monthly downtime for industry

Source : eMaint
 1.4 Maintenance Approach
Functional View.
4. Condition-based Maintenance (1)
1. Reactive Maintenance - Condition monitoring via sensors
- Run-to-failure strategy - Maintenance is performed, only when equiopment problems
have been registered.
- Result : Low routine maintenance cost but high cost
in case of equipment failure and risk of long Results : Anomalies are identified and resolved prior to
downtimes. functional failure.
2. Preventive Maintenance Implementation variations : Hand-held devices or
retrofitted/previously integrated sensors.
- Maintenance is planned and performed in regular
intervals 5. Predictive Maintenance (2)
Without using
- Time or usage triggers are used to schedule - Condition monitoring is enhanced by advanced statistics,
Sensing maintenance. sensing stochastics, real-time analytics or even machine learning
algorithms, which allow to make predictions on when
- Result : Reduced likelihood of failure, but an
equipoment will fail.
ongoing-effort is necessary.
Result : Equipment failure is predictive and preventive actions
3. Proactive Maintenance
can be taken.
- Determining the root causes for machine failure
Implementation variations : Using static-rule-based analytics
- Taking measures/corrective actions to avoid or using dynamic-model based analytics
equipment vailure altogether e.g give workers
6. Prescriptive Maintenance
training on best-practice machine operation.
- Machine ‘self-diagnose” and schedule maintenance
- Result : reason for equipment failure are avoided
- Result: Completely automated maintenance workflow
1. Sometimes also referred to as “diagnostic”
2. Sometimes also referred to as “prognostics” Source : IoT Analytics Research
 1.4.1 Global Survey - Maintenance Programs Adoption
The number of companies that perform predictive maintenance is still less than
the others. Advantages Disadvantages
- Low Cost - Frequent unplanned downtime.
Maintenance Adoption - Less Staff - Increased labor cost, especially
Reactive if over time is needed.
12%
- Possible secondary equipment
Predictive
from equipment failure

- Flexibility allows for the - Unplanned downtime of

adjustment of maintenance equipment
55% periodicity.
Reactive - Unneeded maintenance
Preventive - Increased component life cycle
- Labour intensive
- Estimated 12-18% cost savings
31% over reactive program.
Preventive
- Increase component - Increased investment in
operational life/availability diagnostic equipment
Source : Rolland Berger Strategy Consultants, Survey 2014 - Decrease in equipment or - Increased investment in staff
Predictive process downtime training
- Decrease in cost for parts and
labor
 1.4.2 Global Survey
Industry using Predictive Maintenance
Life Sciences
HIGH
(Pharma, Medical and
Bio Tech)
Oil & Gas

Big Data & Analytics Food &

manufacturing Beverage
opportunity (Based Mining Automotive &
on Volume and Aerospace
applications)

Chemical Others
(industrial equipment, hi-
tech and general
Power Generation manufacturing)

LOW Pulp & Paper

0% 1.5% 3%
IT Spend as a percent of Vertical Revenue Source : Frost & Sullivan 2015
 1.4.3 Global Survey
Industry using analytics applications

Production Line Level Plant Level

HIGH HIGH
System Inventory
Asset
Downtime Management
Performance
Quality
Enterprise/End User

Operational

Enterprise/End User
penetration level

penetration level
Asset Assurance Downtime
Enterprise
Utilization
Visibility
Product Lifecycle
Asset Management
Availability Resource
Allocation Financial
Material Analysis Demand & Supply
Wastage Forecasting
LOW LOW
LOW Opportunity for analytics at HIGH LOW Opportunity for analytics at HIGH
production line level production line level

Source : Frost & Sullivan 2015

 1.4.4 Global Survey - Total PdM Market
Predictive Maintenance a 1,498M USD Market in 2016, expected to grow to
10,962M USD by 2022
Key Drivers
1. New PdM Business Models Global Market Development (PdM)
The Growth of Pdm is to a large extent driven by PdM- 12000
10962
based business models. (e.g., ThyssenKrupp’s MAX
elevator service) 10000
CAGR 39% 8149

Global Market Size

2. Senior Executive Buy-In. 8000
As (industrial) firms double down on their “digitization

in M USD
efforts”, CEOs/CDOs increasingly see the benefits of 5983
6000
IoT/Digitalization/PdM, subsequently push for PdM
4305
projects.
4000 3070
3. Decrease in cost for IoT technology 2154
IoT technology is getting cheaper at a fast pace due to 2000 1498
standardization and scale effects, thereby making the
business case for PdM more attractive. (e.g., declining 0
prices for cloud, connectivity) 2016 2017 2018 2019 2020 2021 2022
Year
4. Favourable Technology Replacement from CBM to
PdM
Some existing condition monitoring solutions are Source : IoT Analytics Research
upgraded to more sophisticated PdM solutions
 1.4.5 Global Survey - Predictive maintenance maturity level
• periodic physical inspections PdM Maturity Level
Level 1 • conclusions are based solely on
inspector’s expertise. 40%
Visual inspections
35%
36%
30%
• periodic inspections;
Level 2 • conclusions are based on a 25% 27%
Instrument combination of inspector’s
20% 22%
expertise and instrument read-outs.
Inspection 15%

10%
• 11%
Level 3 Continuous real-time monitoring of
assets, 5%
Real-time • alerts given based on pre- 0% 3%
conditioning established rules or critical levels. No Pdm Level 1 Level 2 Level 3 Level 4
monitoring
Source : PWC, Predictive Maintenance 4.0 , 2017
• continuous real-time monitoring of
Level 4 assets,
PdM with Big Data • alerts sent based on predictive • 2/3 of respondents are still below level 3 for PdM.
Analytics techniques, such as regression • Only 11% have already reached level 4
analysis.
 1.5 Current Maintenance Technique - Condition Monitoring
Monitor the condition of the machine to know its health.
Key Takeaways Key Takeaways

• The machine is trying to tell its condition via : • Each technology tells something about the machine.
- Vibration • The best way is to select the technologies and the test frequency
carefully.
- Temperature
• Best results are achieved when the results from multiple
- Oil Sample
technologies are integrated into one report.
- Sound
• Some technologies can tell us that a problem exists now, and they
- Motor voltage / current. help to assess the severity.
• There are many technology required to assess the condition. • Some tests will also tell us if a situation exists that could case a
• If we monitor the condition and take appropriate action we can problem :
reduce the risk of catastrophic failure. - Contaminant in the lubricant
• The benefits include : - Resonance
- Reduced downtime - Misalignment
- Reduced labor cost
- Reduced parts inventory • If we can assess the risk of failure we can plan the repair at the
- Greater profit most cost effective time. Parts can be ordered, and labor
organized.
 1.5.1 Vibration Monitoring
Rotating machinery will vibrate. Imbalance

Misalignment
• The patterns and the amplitude tell us about the nature and
severity of any fault conditions.
• We measure the vibration and look at the levels and the
pattern and how they change.
• Vibration can tell us many type of faults :
Flow
Problems Bearing
Broken rotor
Faults Looseness bars

How we do it?
• We mount the sensor on a bearing.
• The sensor and data collector are taken from machine to
machine
 Vibration Analysis
• We can study the time waveform and the
spectrum to learn about the condition of the
machine

Routine Monitoring
• Task are typically performed every 30 days.
• Test a machine at 2 or 3 bearings.
• Collect vertical, horizontal and axial data on the
machine
1.5.2 Infrared Thermography
Thermography is the study of radiated energy using a thermal infrared imaging
system.
Key Takeaways Measuring Devices
• Heat can indicate a fault condition. • Spot ‘radiometers’ -> temperature in an area
• Temperature can be measured at a point, but we will • Inexpensive and expensive cameras
focus on the thermographic image.
‒ Like digital cameras, but infrared image
• Movie cameras
- Record imaging images
• Most come with software for analysis and reporting.
 Infrared thermography Typical maintenance application
• The camera detects radiated energy and display the temperature • Mechanical
gradients. A scale indicates the relationship between color and
- Machines, pipes, bearings, belts
temperature.
• Electrical
• The radiated energy is interpreted by the camera to display a
temperature at a point or in an area. - Overhead lines, transformers, motors, control panels.
- The actual temperature indicated may be incorrect. • Steam system
- Surface type, wind conditions and the other factors affect accuracy. • Refractory plant

The source of heat may come from

inside the object (tank, machine,
cabinet) but the detector only see the
energy emitted from the surface
1.5.3 Oil Analysis
Oil is the life-blood of your rotating machinery
Key Takeaways
• Oil is essential and should be maintained in good
condition.
- The viscosity and other qualities must be OK
- It must not be dirty or contaminated
• Testing the oil and grease reveals :
- If the lubricant is fit for the task
- If it is contaminated, example : water, dirt
- If there are any signs of wear within the machine (metal Oil Analysis Test
particles)
- Oil Bath 40C & 100C → viscosity
• Oil samples are collected routinely.
- R.D.E Spectroscopy → Elemental Concentration
• Samples are either analyzed on-site and/or sent to a
laboratory. - FT-IR (infrared) → Degradation, contamination, additive depletion
• The laboratory will perform a variety of test : - Total Acid → Acid Levels
- Moisture/water content - Total Base
- Viscosity - Crackle → Concentration to 200ppm
- Elemental Analysis - Karl Fisher → Concentration to 10ppm
 1.5.4 Acoustic Emission or airborne ultrasound
Rotating equipment and other assets emit high-frequency sound that provides
clues to potential problems.
What is Ultrasound? How do you use it?

Subsonic Ultrasonic
Sonic Range
Range Range

• The humand ear can only detect sound in the sonic range 20
to 20.000Hz .
• Any sounds that are above the humans hearing capabilities
are referred to as “Ultrasonic”

• An ultrasonic Gun is required to measure the signal and then

heterodyne (demodulate) the signal to an audible signal.
 Some Application Include :
What do you do with it?
Air Leaks

• Air leaks are typically found in the 40khz range.

• Huge cost to industry, a small leak over one year can cost
thousands of dollars.
• Plant noise can hide audible noise.

There are two ways to use the ultrasound data.

• You Can listen to it
- Search for leaks, listen for faults, etc.
• You can measure it.
- The units are displayed in dB
- You can also view a time waveform and specturm
 Faulty Steam Traps Electrical Problems

• A typical failure raite is 30% - 40% of a facility’s steam trap population. • Commonly found in :
• Faults generate ultrasonic turbulence, valve bounce and float - Motor control centers
impacting. - Breaker panels
- Power Lines
- Connections
- Insulation Breakdown
- Improper or Broken
connections
 Bearing Faults & Lubrication Mechanical Fault Detection

• A high percentage of bearing fail due to incorrect, insufficient • Rotating Equipment

lubrication which generate frequencies in the 30khz range. Ultrasonics
• Bearings
can help detect this problem.
• Gearbox
• Bearing impacts can clearly be heard, and dB levels trended.
• Pumps
• You can even lubricate the bearing while listening via the ultrasound
tool. • Motors
• Compressors
WIRELES STEAM TRAP MONITORING
 2.1 Industry Landscape - The use of steam in Industrial Facilities
Steam is used both for heating and for performing various process operations.

• Ammonia is manufactured from Nitrogen (air)

• Water is heated, turns into steam and spins a
and hydrogen (made from natural gas, naptha
steam turbine which drives an electric
or coal with steam)
generator.
Power Generation Fertilizer

• Steam is used to cook wood chips (along with • Pure steam is mainly used for sterilzing tanks,
the cooking chemicals) filters, and piping system, as well as products
• Steam is also used to produce power in sterilizer.
Pulp & Paper Pharmaceutical
• It takes typically 42 kg of steam to refine 1
barrel crude oil. (can vary between 28kg to • Steam injection is an increasingly common
85kg). method of extracting heavy crude oil.
• Steam accounts for over 50% of the total • It is considered as an Enhanced Oil Recovery
energy consumption in a typical refinery. (EOR) method and is the main type of thermal
Refinery • 9415 bpd = 395000 tonnes/day of steam. (as Upstream O&G stimulation of oil reservoir.
reported by EIA).
2.2 Condensate
As the steam gives up heat, it cools down and forms droplets of water called
condensate.

Accumulated condensate in a steam system piping can

cause a number of costly problems

Under low temperature conditions

Burst condensate can freeze and piping
pipes resulting in damage such as burst pipes
and costly shutdowns.

Water Droplets
Water Source : 123RF.com
Hammer
Condensate that builds up in a steam system pipes
Accumulated condensate can also cause water hammer which is a Efficiency also impedes the transfer of heat from the steam
condition that occurs when condensate is carried along with the and takes up space that could otherwise be
reduction occupied by steam. Both of these actions reduce
steam inside a system piping and slams into the wall of a pipe
particularly around bends. This action can break or crack the pipe. the efficiency of the steam system.
2.3 Steam Trap
Is a device used to discharge condensates without allowing valuable steam to
espace.
Key Fact
- A device that’s comonly used for this purpose is a steam
trap.
- Steam is a vapord that’s formed from water.
- The condensate that’s formed as steam cools is a liquid
that’s heavier than the steam.
- In a steam system, the relatively heavy condensate will
drain to the system low points.
- So steam traps are usually installed at these locations to
remove the condensate.
Steam Trap, Source : dreamstime.com

Thermostatic Thermodynamic (TD) Float & Thermostatic (F&T) Inverted Bucket (IB)
It sense temperature change of Operate on the velocity Uses difference in density Uses difference in density
condensate to operate valve principle between steam and condensate between steam and condensate
 2.3.1 Steam Trap Locations - Tank Farms
generally have a large number of steam traps installed

Source : Spirax Sarco

Tank Farms,
Source : 3D line locating.com • Often have other wireless devices installed monitoring
pressure, temperature and level.
 2.3.2 Steam Trap Locations - Flare Stacks
generally have a large number of steam traps installed as part of steam feed to
nozzle manifold.

Flare Stacks, Source : youtube.com

• It has to operate within strict guidelines to meet environmental

legislation
Source :spirax sarco
2.3.3 Steam Trap Location - Steam Turbines
If the traps fail to close and condensate enters the turbine there is a huge risk
that turbine blade damage can occur.

Steam turbine blade, Source : Power Magazine

Steam trap in steam turbine, Source : Youtube.com

 2.4 The need to monitor Steam Trap
Steam traps performance now seen as the key to improve plant efficiency
• Steam leaks from traps are costly in both financial and
environmental sense and therefore need prompt attention to
Steam ensure that a steam system is working at its optimum efficiency.
Leaks • With the reduction in upstream investment many end users are
now investigating ways to improve process efficiency
downstream

• A global focus on carbon emissions has resulted in many

Carbon
organisations being more proactive in reducing their energy
Emission usage as part of wider environmental policy

Loss of • Blocked or failed-closed steam traps, if ignored, can lead to

production the eventual failure of the entire asset damaging equipment
or pipeline infrastructure.
• Approximately 650.000 process traps installed in time
refineries worldwide.
• Blocked stream traps have caused the solidification of entire
• Typically 10 – 15% will fail each year.
product lines, such as suphur. In this case the whole affected
Solidification section require removal and replacement, shutting down the
process, and lost production output.
2.5 Current Monitoring Technique
How plant usually monitor the steam traps?
Painpoints
• In many companies, steam traps are checked
manually by a specially traned engineer listening to
each individually on an anual basis.
• Due to the large trap population on many oil and gas
facilites there is a potential that failed traps could go
unnoticed fo an extended period of time.
• This increases the potential for unexpected
equipment or pipeline failure leading to significant
energy loss or more importantly lengthy process
downtime.

Source : Synergys Technologies

Energy Loss Process Downtime

 2.6 Example Solution
Emerson Steam trap monitoring system

Wireless Acoustic Transmitters , Source : Emerson

• Intelligent monitoring also allows maintenance to prioritise critical

steam traps during scheduled downtime ensuring high priority
processes are protected from potentially avoidable interruptions.
• Removes the need to send manpower in to potentially hazardous
process area.

Plant Web, Source : Emerson

Thankyou