Transcript Slide 1

HA-PI at Northparkes
Mines
HA-PI: Context
• Where we use PI
• Why we use PI
• How we use PI
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
2
Location
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
3
Northparkes Company
• A series of porphyry copper-gold deposits with
proven resources until 2017.
• Mining began in 1993 with open pit oxide gold
• The current method of mining, block caving, was
commenced in 1998. 5.5Mtpa is processed at
1.5% Cu head grade.
• Rio Tinto bought North mining in 2001
• Concentrate is sold to various smelters in Asia –
with 20% going to our joint venture partners
Sumitomo.
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
4
Northparkes flow sheet
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
5
UG mine layout
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
6
Block Cave Terminology

Undercut block by blasting and
removing a slice of ore

Remaining ore fractures along
jointing under induced stresses

Broken ore is removed through
drawbells and trammed on
extraction level to crusher
Undercut
Draw Bell
Extraction Level
Access Decline
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
7
Flotation separation
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
8
Background with OSIsoft
• PI was installed in November 2006 – no historian
had been used before.
• A 2000 tag system was purchased, covering all
of the current trend tags on the Citect system
(mill) and the reporting tags from the mine (a
MacroView system).
• Currently only the metallurgy team directly
access the data.
• Data is accessed through processBook and
datalink (10 concurrent licenses of each).
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
9
Why PI?
• PI was purchased to align NPM with other Rio Tinto
users, particularly Rio Tinto Aluminum and Pilbara Iron
• NPM have moved to the Rio Tinto Corporate domain – PI
was the first server on site to run on this domain.
• Ad hoc investigations are the major use at this time.
• Enable centralized monitoring and support of process
control loops. Similar work is ongoing at Pilbara Iron,
already contributing over $100m of value.
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
10
Why high availability?
• We plan to have PI at the heart of our reporting
systems.
• HA PI, combined with Live State and normal
corporate backups allows for greater reliability.
• Fully redundant system has been tested, and
after recent power failures restarted
automatically.
– Allows IT to patch individual components without loss
of data.
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
11
Plans for the Future
• PI will become the only process data source for
corporate applications as they are upgraded or
replaced.
• Capture metallurgical knowledge in the form of
processBook display documents, held in a
structured directory.
• Restructure data flows.
• Allow viewing of mine information at “mill control”
and vice versa.
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
12
Old data flows
UG PLCs
NPMPROD
OPD PLCs
Macroview
Citect i/o server
OPC server
Matrikon
product server
PGMNPM
Lab data base
Elipse
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
users
13
Primary PI Server
PI Collective
Pi Clients
CORP.RIOTINTO.ORG LAN
Mill
Router
OPC Server
OPC Server
Pi Data Collector
OPC Server Failover
Pi Data Collector
Data collector Failover
NPMPROD LAN
Mine
OPC Servers & Pi Data Collector
N-Way Buffering
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
14
Proposed data
flows
OPD PLCs
UG PLCs
Macroview
OPC servers with
PI interface nodes
Citect i/o servers
PI interface
nodes
NPMPROD
PI historian
Elipse/SAP
Matrikon
product server
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
Lab data base
CORP
users
15
Conclusions
• System alignment within Rio Tinto a key
driver toward PI.
• HA PI has been implemented.
• NPM challenges are
– Extract value from the data
– Capture metallurgical experience in the form
of relevant process books
– Improve the data flows on site.
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
16
Questions
Copyright © 2007 OSIsoft, Inc. Company Confidential. All rights reserved.
17