“THE POWER OF 1%” (PLN Turbine System) BIG DATA ASSIGNMENT

“THE POWER OF 1%”

(PLN Turbine System)

BIG DATA

by Wirawan Rizkika 1401140469

 

First, i choose this system because of this system is common in our daily life, which is Distributing Electricity, and most importantly, it is needed to supply our electricity to every region of Indonesian islands. PLN (Perusahaan Listrik Negara, English: ‘State Electricity Company’) is an Indonesian government-owned corporation which has a monopoly on electricity distribution in Indonesia. Literally, as PLN is the only one company which distribute electricity in Indonesia, PLN has many Turbines or Generators to generate the electricity. To operate those massive turbines across Indonesian Islands, PLN needs systems that are integrated to perform its functions. And having those equipments also need maintenance to keep the system working properly, the system also enable PLN to predict its maintenance.

 

Operations

 

In the first half of 2011, the PLN generated 88 terawatt-hours (TWh). The firm generated around 24% of its output using oil-based fuel with plans to reduce the share to 3% by 2013 and 1.7% by 2014. The forecast for the full year (2011) is around 182 tWh (equivalent to around 760 kWh per capita).

 

Capacity and organisation

At the end of 2011, the PLN’s total generating capacity (produced by a many different plants across Indonesia) was estimated at around 28,500 MW. In 2012, a combined capacity of 3,351 megawatts will come online from 23 new power plants.

 

PLN: Capacity and peak load, end-2011 (megawatts)

Maximum capacity Peak load
Java-Bali 21,257 16,150
Western Indonesia 4,602 4,299
Eastern Indonesia 2,603 2,484
Total 28,462 22,933

 

Based on data above, PLN will surely needs proper and effective system in managing the equipments, products, etc.

COMPONENT IN THE SYSTEM

How does the system works?

pln-system-2

Micro services Involved:

  • Interconnecting Environment:
  1. Edge Manager (Connecting external cloud to cloud foundry)
  2. Machine (Connecting Turbines with cloud foundry)

 

  • Data Management:
  1. Time Series (Real-time data streaming)
  2. SQL Database (Real-time assets reporting purposes)
  3. Asset Data & UAA (Registering turbines and assign privileges)
  4. Blob Store (Archiving historical data)
  • Analytics
  1. Anomaly Detection (Predictive and Preventive analytics for each turbines)
  2. Analytics User Interface (End user UI)
  • Operations
  1. Logging (Operational System up-time monitoring)

 

Purpose of the system:

  1. Save supply
  2. Healthy environment
  3. Asset Performance Management (Predictive Analysis)
  4. Increase up-time and productivity

 

Communication between system uses Rest API.

Every asset connected to the cloud are acting as Slave Node. All inside the cloud foundry environment are acting as Main Node.

PROCESS EXPLANATION

Archive : Aggregated Historical Archive are used to save operational cost. Assets Performance Management (APM) will gather some aggregated data + alarm code from Historical Data Archive, to be compared with real time data, which will results in Predictive Maintenance. (So, PLN will be notified if there’s some parts are going to be wear out or broken soon, and need to be replaced. PLN will have time to prepare for its maintenance. PLN will loss 2 million dollar per day if the Turbines are broken. Cost to have the license of the software of each turbine is 50.000$ a year.)

The turbine will generate electricity, on each turbine, it is equipped with high technology sensors, and the PLC of each turbine will read each parameters or indicators in the turbine. Those data will be transferred and processed inside the cloud foundry by the micro services. Each micro services has its specific function. Each data are secured by multiple level of security layer, such as the Wurldtech Device that has developed algorithm to allow or block certain communications to or from any kind of protocols.  The micro services itself is able to work in parallel or in sequence depending on the configuration set. Two ways communication is possible between the Slave Node and the Main Node. 95% of data will be going from Slave Node to Main Node, 5% of the data will mostly be in the form of commands. From all assets data by using the APM Micro services, it will form a historical algorithm that will be able to detect any abnormality until the most detailed level, e.g Turbine A will need to have a gear maintenance within the next three months due to the number of alarm events triggered in the past. The APM  will also give a notification of which part should be replaced or maintained. This will require an investment of 50.000 USD a year, hence it will minimize the risks of two million dollars for each turbine a day in case of downtime.

One of the biggest benefits the system can produces or deliver to the customer is reducing the gas or oil supply for each turbine. The APM will tell the PLC to work at the very optimized rate. Which means the turbines doesn’t have to fully works at 100% productivity in case of 75% demand.

Predix Illustration :

home_diagrame_2x

Length of Time:

Basically, this system will work 24 hours a day, 7 days a week. PLC Sensor generates data every 2 seconds. Every 300GB data will be erased or moved to Historical Data Lake.

Number of suggestions related to points in “Strategy in Managing Big Data Analytics”:

  • Best Tools
    • By applying best tools (5), the process can be managed by cloud computing. Cloud computing creates high flexibility and does not depend on any platform. Therefore by using cloud computing in any media, regardless of the platform, we will still be able to access the same. Thus, cloud computing is more centralized.
  • Data Share (4)
    • By utilizing Data Share (4), the process can be shortened. Rather than manually checking each turbines, each turbines has its system to manages and even predicts the maintenance, or even can be controlled from the Main Hub.
  • Scalability and speed (6)
    • By utilizing Scalability and Speed (6), the process will be more effective and simpler. Scalability and Speed allow the system to cope with multiple turbines in an effective time.

Wirawan Rizkika 1401140469

 

Sources :

PLN Definition and Capacity : Wikipedia.com

Analysis and Diagram made by Wirawan Rizkika

Predix.io

Internal GE Solution Architect

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s