Deckblatt

Prevention of overload using

"CPU_RT"
SIMATIC PCS 7 V7.0 and higher

Application Example y November 2011

Applications & Tools

Answers for industry.
 Industry Automation and Drive Technologies Service & Support Portal

This article is taken from the Service Portal of Siemens AG, Industry Automation
and Drive Technologies. The following link takes you directly to the download page
of this document:
http://support.automation.siemens.com/WW/view/en/55692652

If you have any questions about this document, please contact us under the
following e-mail address:
[email protected]

For further information on this topic, you may also actively use our Technical
Forum in the Service & Support Portal. Add your questions, suggestions and
Copyright © Siemens AG 2011 All rights reserved

problems and discuss them in our large forum community:

http://www.siemens.com/forum-applications

CPU_RT
2 V 3.0, Entry ID: 55692652
 Task Description and
1
s Solution

Configuration and
Settings 2

Summary 3

Links & Literature 4

SIMATIC PCS 7
History 5
CPU_RT

Application Example
Copyright © Siemens AG 2011 All rights reserved

CPU_RT
V 3.0, Entry ID: 55692652 3
 Warranty and Liability

Warranty and Liability

Note The application examples are not binding and do not claim to be complete with
regard to configuration, equipment or any contingencies. The application
examples do not represent customer-specific solutions; they are only intended to
provide support for typical applications. You are solely responsible for the correct
operation of the described products. These application examples do not relieve
you of your responsibility to use sound practices in application, installation,
operation and maintenance. When using these application examples, you
recognize that we will not be liable for any damage/claims beyond the liability
clause described. We reserve the right to make changes to these application
examples at any time without prior notice. If there are any deviations between
the recommendations provided in this application example and other Siemens
publications (e.g. catalogs), the contents of the other documents shall have
priority.

We do not accept any liability for the information contained in this document.
Any claims against us – based on whatever legal reason – resulting from the use of
the examples, information, programs, engineering and performance data etc.,
described in this Application Example shall be excluded. Such an exclusion shall
Copyright © Siemens AG 2011 All rights reserved

not apply in the case of mandatory liability, e.g. under the German Product Liability
Act (“Produkthaftungsgesetz”), in case of intent, gross negligence, or injury of life,
body or health, guarantee for the quality of a product, fraudulent concealment of a
deficiency or violation of fundamental contractual obligations. Damages for a
breach of a substantial contractual obligation are, however, limited to the
foreseeable damage, typical for the type of contract, except in the event of intent or
gross negligence or injury to life, body or health. The above provisions do not imply
a change in the burden of proof to your detriment.

It is not permissible to transfer or copy these Application Examples or excerpts

thereof without express authorization from Siemens Industry Sector.

CPU_RT
4 V 3.0, Entry ID: 55692652
 Table of Contents

Table of Contents

Warranty and Liability ................................................................................................. 4

1 Task Description and Solution......................................................................... 6

1.1 Function................................................................................................ 6
1.2 Solution ................................................................................................ 6
1.2.1 Prerequisites and validity ..................................................................... 6
1.2.2 Functional scope .................................................................................. 8
1.2.3 Applications .......................................................................................... 9
2 Configuration and Settings ............................................................................ 10
2.1 New controller overload behavior....................................................... 10
2.1.1 Configuration of cycle overload behavior ........................................... 11
2.1.2 Signaling behavior with overload ....................................................... 14
2.1.3 Cancelation of emergency mode in case of controller overload ........ 14
2.2 Cycle Performance Analysis .............................................................. 15
2.2.1 Display of cycle performance data on OS/MS ................................... 15
2.2.2 Scenario 1: High controller cycle load (maintenance demand).......... 18
2.2.3 Scenario 2: OB request error (maintenance request) ........................ 19
2.2.4 Scenario 3: Exceedance of the maximum cycle time
Copyright © Siemens AG 2011 All rights reserved

(maintenance alarm) .......................................................................... 20

3 Summary .......................................................................................................... 21

4 Links & Literature ............................................................................................ 22

5 History............................................................................................................... 22

CPU_RT
V 3.0, Entry ID: 55692652 5
 1 Task Description and Solution
1.1 Function

1 Task Description and Solution

1.1 Function

The controller of an automation system has two primary tasks. On the one hand
the controller automates the process and on the other hand the controller has to
evaluate and forward any operations and displays.
All tasks are processed sequentially in the controller (one after the other). In order
to ensure an appropriate processing time, the controllers of the SIMATIC product
line have a basic cycle monitoring (OB1) of 6 seconds maximum. This leads to a
time limitation of applications, in addition to the memory-based limitation. If the
cycle checkpoint is not reached within 6 seconds, OB80 (time error) will be called
and a system message will be issued.
For process automation, SIMATIC PCS 7 provides the option of reacting to
temporary and permanent cycle overload through controlled, automatic load
shedding. This functionality is realized by the "CPU_RT" function block and will be
available as of PCS 7 version 7.0.

1.2 Solution
Copyright © Siemens AG 2011 All rights reserved

The required load relieving measures in case of a PCS 7 controller overload are
automatically performed by the "CPU_RT" function block“. The "CPU_RT" function
block combines the following functions:

• Maintaining operability through a new overload behavior by means of multi-

stage reduction of the cyclic load.

• Prevention of overload through a cycle performance analysis and visualization

of performance data in the Asset Management

1.2.1 Prerequisites and validity

SIMATIC PCS 7 version V7.0 or higher is required for the automation solution.
The function for determining the cycle performance data of the "CPU_RT" function
block requires a controller firmware revision level that supports the "SFC 78"
system function.

Table 1-1: "SCF 78" support for "SFC_78"

Controller Firmware versions
for single controller 5.0 or higher
for H controller 4.5 or higher

CPU_RT
6 V 3.0, Entry ID: 55692652
 1 Task Description and Solution
1.2 Solution

If you compile CFCs with the option "create module driver", a plan named
"@CPU_RT" will be created as of PCS 7 V7.0. It will include the "CPU_RT"
function block, which will be integrated in all relevant execution levels (OB100,
OB1, all OB 3x, and OB 8x) by the module driver wizard.

Figure 1-1: "CPU_RT" block

Copyright © Siemens AG 2011 All rights reserved

Note The "CPU_RT" block is placed into the system plans by the PCS 7 module driver
wizard. All controllers will feature the new overload behavior as of PCS 7 V7.0 in
combination with the V7 Library.

When an older project is migrated without utilization of new functions, the driver
wizard will not integrate the "CPU_RT" in PCS 7 V7.0, either.

Apart from the "SFC 78" system function the "Asset Management" option package
is required for full utilization of the new controller overload behavior and the cycle
performance analysis.

CPU_RT
V 3.0, Entry ID: 55692652 7
 1 Task Description and Solution
1.2 Solution

1.2.2 Functional scope

Depending on the CPU/FW used and whether or not Asset Management is

implemented, the functions of CPU_RT will be displayed and parameterized as
follows:

Table 1-2: Overview functional scope

Case 1 Case 2 Case 3

Controller FW Controller FW Controller FW

Function Constellation with SFC 78 and without SFC 78 without SFC 78
option package and option and without Asset
Asset package Asset Management
Management Management
available available
Cycle performance analysis YES NO NO
Display of cycle performance
YES NO NO
data on OS
Display of cycle performance
Copyright © Siemens AG 2011 All rights reserved

YES NO NO
data in the CFC
Messages YES YES* YES*
Parameterization via NO
faceplate YES YES
(via CFC)
New controller overload
YES YES YES
behavior

* No operator message "Net time requirement of all OBs exceeds Max Limit", since
performance data cannot be calculated.

Note The function of the "CPU_RT" block is described in detail in the manual
"SIMATIC Process Control System 7 PCS 7 Basis Library V7.1".

CPU_RT
8 V 3.0, Entry ID: 55692652
 1 Task Description and Solution
1.2 Solution

1.2.3 Applications

The following applications were taken into account in the development of the
"CPU_RT" functional block:

• Applications with little time reserve have the highest risk of overload. Systems
with high time load can go into overload quickly through a defective
component, typically a defective measuring line that produces an alarm flood
(chattering alarm). The "CPU_RT" block enables you to determine the overall
runtime (with interruptions by higher-priority process levels) through OBs
during commissioning and in normal operation, apart from the net runtime of
the organization block (OB). Early load relief of time-critical OBs represents a
good protection against controller overload.

• Changes and extensions of the user program with an already high cycle load is
another application. Changing or extending the user programm causes an
overload of the controller after loading into the automation system (AS). The
cycle overload of the controller is selectively reduced through the "CPU_RT"
block. Process levels or OBs representing a very high load are temporarily
reduced (load shedding - emergency operation) until the temporary load has
Copyright © Siemens AG 2011 All rights reserved

diminished. This ensures permanent operability of the plant even in case of

overload. It is possible to reverse changes during operation or to restructure
the user program accordingly, in order to return to normal operation. One
example for the case described is load reduction by transferring program parts
to slower OBs.

• Adding communication lines may lead to a controller overload under certain

circumstances. In this case, too, the block allows for cancelation of the change
in "emergency" operation.

If an overload occurs during operation for no obvious reason, you can localize the
cause by a detailed performance analysis of the "CPU_RT" block. In particular the
net times (net OB runtime without interruptions) must include an indication as to
where additional time losses have occurred (see chapter 2.2) This can even be
done retroactively in normal operation, due to the new overload behavior.

CPU_RT
V 3.0, Entry ID: 55692652 9
 2 Configuration and Settings
2.1 New controller overload behavior

2 Configuration and Settings

2.1 New controller overload behavior

In case of a controller cycle overload you can prevent the controller from being
unoperable through controlled load shedding.
Load shedding is effected by the "CPU_RT" block and is achieved by reducing
cyclical process levels. Load shedding represents an emergency mode, the cause
of which needs to be eliminated as soon as possible.
Emergency mode is initiated when the cycle checkpoint monitoring time has been
exceeded twice. In this case the cycle checkpoint has not been reached for at least
12 seconds due to a temporary or permanent overload.
The process control message "emergency mode, cyclic OBs are reduced" is issued
with standard priority "16". Therefore this message is displayed in the list of
highest-priority messages.

Note With the "Asset Management" option package available an additional

maintenance alarm will be generated at the Asset faceplate of the CPU
concerned.
Copyright © Siemens AG 2011 All rights reserved

Figure 2-1: Emergency mode of the controller

CPU_RT
10 V 3.0, Entry ID: 55692652
 2 Configuration and Settings
2.1 New controller overload behavior

In emergency mode the blocks are skipped x times, independent of the reduction
factor, so that you can continue processing all blocks.
The following two escalation levels are available:

• Level 1:
The cycle checkpoint monitoring time has elapsed twice (approx. 12s). The
OBs 3x released for reduction are reduced once.

• Level 2:
If time errors continue to occur, all OB 3x are incrementally reduced.

The emergency mode is not meant to be a permanent bridging of the "cycle

Note
overload" state of the controller, but is rather used for temporarily handling
overload situations.
The cause for the emergency mode must be eliminated without delay. This is
also the reason why the emergency mode will be reported as a maintenance
alarm in the Asset Management.
In order to change back to normal operation you can, for example, restructure
the program or reset / transfer the changes made.
Copyright © Siemens AG 2011 All rights reserved

2.1.1 Configuration of cycle overload behavior

If the Asset Management option package is available, the required overload

behavior is parameterized in the Asset Faceplate of the CPU,
by the following procedure:

1. Start the OS runtime.

2. Open the Asset Faceplate of the CPU.
3. Parameterize the desired overload behavior.

In Figure 2-2 below, OB 33 and OB 35 are excluded from the first escalation level
and are not reduced there.

CPU_RT
V 3.0, Entry ID: 55692652 11
 2 Configuration and Settings
2.1 New controller overload behavior

Figure 2-2: Parameterizing the block "CPU_RT"

Copyright © Siemens AG 2011 All rights reserved

CPU_RT
12 V 3.0, Entry ID: 55692652
 2 Configuration and Settings
2.1 New controller overload behavior

Note In case the Asset Management option package is not available to you, you will
have to parameterize the overload behavior directly at the "CPU_RT" block in the
CFC plan "@CPU_RT". This procedure is described in following.

The CFC plan "@CPU_RT" is a system plan and must not be deleted.

Reduction factor (max. number of retrigger attempts)

The overload behavior is parameterized at the "MAX_RTG" input. Here you can set
the maximum number of retrigger attempts.

Note If you assign "0" to the "CPU_RT" input, the overload behavior will be as in older
SIMATIC PCS 7 versions (< V7.0).

• MAX_RTRG=1
Copyright © Siemens AG 2011 All rights reserved

Only the first level of the new overload behavior is initiated.

If the load shedding through level 1 is not sufficient, the controller will go to
STOP at "MAX_RTRG" = 1 after the cycle checkpoint monitoring time (>18s)
has been exceeded three times, and level 2 will not be initiated.

• MAX_RTRG = 2
After the cycle checkpoint monitoring time (18s) has been exceeded three
times, level 2 will be initiated. All OBs will be reduced once. If this load
shedding is not sufficient, either, the controller will go to STOP state after the
cycle checkpoint monitoring time (>24) has been exceeded four times.

• MAX_RTRG = x
If you assign a value > "2" to the input, the cycle checkpoint monitoring time
can be exceeded x times. The reduction factor will be increased with each
exceedance.

CPU_RT
V 3.0, Entry ID: 55692652 13
 2 Configuration and Settings
2.1 New controller overload behavior

2.1.2 Signaling behavior with overload

Overall load alarm limit (only for controller with SFC 78):

If the overall load (sum of mean values of all OB net runtimes) exceeds the upper
limit set at the parameter overall load alarm limit or MAX_VAL (CPU_RT), the
message "net time requirement of all OBs exceeds Max Limit" is output.

Note See chapter 2.2.2 “Scenario 1: High controller cycle load (maintenance demand)

Max. number of OB request errors

If the number of OB request errors parameterizable at parameter "signaling from

request error" or N_REQ_ERR (CPU_RT) is exceeded, the message "OB 3x
request error" will be output.
Copyright © Siemens AG 2011 All rights reserved

Note See chapter 2.2.3 “Scenario 2: OB request error (maintenance request)

2.1.3 Cancelation of emergency mode in case of controller overload

You can automatically end the emergency mode if you reduce the maximum
overall load of the controller configured at the parameter "cancel reduction at" or
MAX_VAL (CPU_RT) correspondingly and no cycle time exceedance occurs on a
permanent basis.
The reductions are also canceled if the slowest OB processes a number of cycles
(Parameter: UNDO_CYC) again.
In this case the value of UNDO_CYC must not be too small in order to avoid
frequent switching between stop prevention measures and normal operation.

CPU_RT
14 V 3.0, Entry ID: 55692652
 2 Configuration and Settings
2.2 Cycle Performance Analysis

2.2 Cycle Performance Analysis

The "CPU_RT" block determines the runtime of the individual OBs and their
participation in the cycle time (cycle performance analysis). This performance data
and load information can help you even in the project engineering phase for
structuring the user program, allowing you to selectively plan reserves for
extensions, utilize the capacity of the individual process level homogeneously, and
prevent later controller overload.

Note The cycle performance analysis is available for controllers with SFC 78 only.

2.2.1 Display of cycle performance data on OS/MS

Preconditions

The cycle performance data is visualized on the OS/MS if the PCS 7 option
Copyright © Siemens AG 2011 All rights reserved

package "Asset Management" is installed.

Note The cycle performance data can only be displayed on OS/MS for controllers with
SFC 78 and Asset Management.

Configuration

Via the Asset faceplate of the controller you can select the faceplate representing
the overall load of the controller (Figure 2-3).

1. Start the OS runtime.

2. Open the faceplate of the corresponding CPU.
3. Select "performance" in the drop-down list.

CPU_RT
V 3.0, Entry ID: 55692652 15
 2 Configuration and Settings
2.2 Cycle Performance Analysis

Figure 2-3: Overall load of the controller

Note The net runtime is the runtime that was needed for processing the code placed in
Copyright © Siemens AG 2011 All rights reserved

OB3x, if it is not interrupted by higher-priority OBs.

The overall runtime is the runtime that was actually needed for processing the
code placed in OB3x, including the runtimes of higher-priority OBs.

The vertical bar represents the overall load of the controller in % and the sum of
all net runtimes (cycle times) of all OBs.

For detail analysis, a further Asset faceplate represents the determined cycle times
in the individual process levels (OBs) (Figure 2-4).

CPU_RT
16 V 3.0, Entry ID: 55692652
 2 Configuration and Settings
2.2 Cycle Performance Analysis

Figure 2-4: OB cycle times

Note The overall load of a single controller after commissioning should not be higher
than 75% (net).
Copyright © Siemens AG 2011 All rights reserved

If 20 percent of the cycle time are used for communication and acyclic events, a
reserve of 5 to 10% will remain for future program extensions or additional
communication orders.

Example

The net runtime of OB38 is 5ms. With a sampling time of 10ms the overall load rate
is 50 percent. In this case the overall and net runtime are almost identical for
OB38, if it is not interrupted by higher-priority OBs.
The net runtime of OB35 is 45ms. With a sampling time of 100ms the overall load
rate is 45 percent. The overall runtime is clearly higher than the net runtime, the
reason being that OB35 is interrupted by OB38.
There is a risk of overload with an overall load of 95 percent. The overall runtime of
OB35 indicates that this OB is close to an OB request error. In case of acyclic
events or high communication load a controller overload might occur in this
described example. Due to the overall load and the overall runtime of OB35 the
program must be restructured.
Program extensions in OB35 or in faster OBs may lead to overload. Here, as a
consequence, you should transfer existing or newly added program parts to slow
OBs.

CPU_RT
V 3.0, Entry ID: 55692652 17
 2 Configuration and Settings
2.2 Cycle Performance Analysis

Conclusion

The program logic and processing for slower process signals (e.g. temperatures)
are typically placed in OB3x with a call interval of 1 to 2 seconds, and dynamic
process values (e.g. print) are processed in faster OBs. This preventive measure is
the decisive measure for preventing a cycle overload of the controller.

2.2.2 Scenario 1: High controller cycle load (maintenance demand)

If the overall load (sum of mean values of all net runtimes in % of OB3x, OB8x +
OB1) exceeds the overall load alarm limit (Figure 2-5), the message "net time
requirement of all OBs exceeds Max Limit" is output and the maintenance status is
set to "maintenance demand".
Due to this preventive maintenance message you must take action for cycle relief
in the user program in order to avoid a cycle overload of the controller, such as
adapting, dividing, or transferring the user program.

Figure 2-5: Net time requirement of all OBs exceeds Max Limit
Copyright © Siemens AG 2011 All rights reserved

Note If "SFC 78" is not supported, the controllers will not signal the exceedance of the
net time requirement.

CPU_RT
18 V 3.0, Entry ID: 55692652
 2 Configuration and Settings
2.2 Cycle Performance Analysis

2.2.3 Scenario 2: OB request error (maintenance request)

If the overall load of the controller reaches the limit of 100%, OB request errors will
be triggered. In this case cyclic OBs are no longer processed appropriately and
called within the specified time frame, since they are, for example, interrupted by
acyclic OBs (diagnosis alarms). This lets the program processing time (overall
time) be longer than the call interval of the OB (e.g. OB35: 100ms).
If the parameterizable number of request errors is exceeded (Figure 2-6), the
message "OB 3x request error" will be issued and the maintenance status is set to
"uncertain maintenance request".
When there is no longer an overload situation, the maintenance status will be reset
and the process control message will be marked as "going out".

Figure 2-6: OB request error

Copyright © Siemens AG 2011 All rights reserved

CPU_RT
V 3.0, Entry ID: 55692652 19
 2 Configuration and Settings
2.2 Cycle Performance Analysis

2.2.4 Scenario 3: Exceedance of the maximum cycle time (maintenance

alarm)

If the cycle checkpoint monitoring time set (HW Config > CPU properties) is
exceeded (PCS 7 default: 6000ms), the message "cycle time exceeded: 6001ms,
OB1" will be output and the maintenance status of the CPU will be set to "bad or
maintenance alarm" (Figure 2-7)
The following events can lead to a cycle overload:

• high communication load

• Alarm flooding

• Frequent acyclic fault events (chattering alarms)

• Misdimensioning of cyclic interrupts when adding new program parts

Figure 2-7: Exceedance of cycle time

Copyright © Siemens AG 2011 All rights reserved

CPU_RT
20 V 3.0, Entry ID: 55692652
 3 Summary
2.2 Cycle Performance Analysis

3 Summary

If you use the new PCS 7 V7.0 Library, the "CPU_RT" block will always be
integrated by the driver wizard. This is done indendently of the use of Asset
Management and the controller firmware.
After an upgrade of the driver components from legacy projects to PCS 7 V7.0 the
controller will show a different cycle overload behavior.
In emergency mode of the controller the cyclic OBs are reduced. This behavior
affects integral computations, fast switch-off functions, or fast control procedures.
In case of overload the STOP controller or the emergency mode controller
(permanent operability) can be used.
Therefore you can set the desired behaviour for cycle overload either at the Asset
faceplate of the CPU or, if the "Asset Management" option package is not
available, directly at the "CPU_RT" block.

Note You can disable the "CPU_RT" block by setting the "MAX_RTRG" parameter
to "0".
Copyright © Siemens AG 2011 All rights reserved

Since controller overload mostly occurs due to misconfigurations or the cumulation

of acyclic events (nuisance /diagnosis alarms OB8x), the permanent operability of
the system and the option of searching and eliminating the error with the system
running is preferred. For this reason the new overload behavior of the controller is
preset as of version PCS 7 V7.0.

CPU_RT
V 3.0, Entry ID: 55692652 21
 4 Links & Literature

4 Links & Literature

The following table gives references to further information.

Table 4-1 Internet links

Topic Title
\1\ Reference to this http://support.automation.siemens.com/WW/view/en/55692652
entry
\2\ Siemens Online http://support.automation.siemens.com
Support
\3\

5 History
Copyright © Siemens AG 2011 All rights reserved

Table 5-1 History

Version Date Revisions
V1.0 12.01.2007 First edition on the DVD PCS 7 V7.0
V2.0 05.11.2009 Layout adaptation for publishing in the Service & Support
Portal under "Applications & Tools" ( see link in Table 4-1)
V3.0 14.11.2011 Überarbeitet; neue Screenshots; Layout angepasst

CPU_RT
22 V 3.0, Entry ID: 55692652