Sei sulla pagina 1di 10

Better Wind Turbine Performance

through Continuous Training of


Operations and Maintenance
Personnel.
Jose San Leandro Ros, jose.sanleandro@acm-sl.com +34605876613

Abstract: The increment in performance in Wind Farms and the reduction of exploitation costs
require inner knowledge of the independent production plants: the Wind Turbines. Developments
on the design of Real Time Simulators have made possible to actually simulate the behavior of
Wind Turbines as a whole plants, to the degree of being able to observe the evolution of internal
signals on normal, exceptional and failure conditions. This technique allows for the training of
Operations and Maintenance Personnel, with a number of advantages related to security,
completeness and, needless to say, costs.

The search for Operational Costs reduction passes by the collection of complete sets of
operational data from the WTs, its study and analysis trying to discover of cross relationships
between them, eventually leading to a better knowledge of the real status of the installations. A
closer knowledge of the WT subsystems is mandatory to perform the fine analysis of the data
gathered in the field: data analysts must have the most complete understanding of all the Wind
Turbine elements, in order to extract the most knowledge out of the (expensive) data.

These findings also must be spread out along all the organization levels (management,
maintenance personnel, etc.), to be truly beneficial.

Unfortunately, understanding the behavior of all elements in a WT, normally, is not an easy task,
even for Maintenance personnel that works close to the equipment. To complicate further the
learning path, the most important and interesting cases occur under abnormal conditions (very
high winds, fast wind gusts, high vibration in the tower, etc), or when a defect arises (reduction on
the Pitch mechanism performance, broken yaw motor, high temperature in the electric generator,
etc.) . But in those occasions, because security reasons, is better not to be nearby them.

So far, the main source for training of Wind Farms personnel and of companies that provide
Maintenance Services, have been training at the OEM premises, complemented with actual
practice in real cases. The training at the OEM premises is “a must” for Certified Maintainers, at
every particular Wind Turbine model. Ideally, these training classes should present, somehow, the
behavior of each of the subsystems in the Wind Turbine, independently as well as working
cooperatively, as in a Wind Turbine in production. Unfortunately showing behavior for a, say, 2
MW Wind Turbine is a risky job inside a classroom.
The experienced maintainer must know the configuration of the subsystems, their specific
functions and possible failures, etc., and how the Wind Turbine’s Controller reports incidences,
alarms or error conditions. For example, if there is a problem with the pump of the brake’s
hydraulic subsystem, failing due to loss of compression power, the Controller will detect it. The
maintainer’s source of information for that problem will be the incidences messages issued by
that Controller, and those messages will be of the type “Pump starting too often”: the experienced
maintainer must know by heart that this type of message may come from a failing pump or
because pressure leakage in the hydraulic circuit. And , before jumping to the Nacelle, at 80
meters high. It is mandatory that the trainees “get used to” all these messages before confronting
a real problem. Memorizing all these messages from reports or written manuals is a painful way of
learning.

It is important to keep in mind that even simple practical knowledge, such as “how long a cutin
process takes to complete”, from stopped drive train to actual production, an operation that is
dependent on the existing wind, may take more observations that the trainee may have
opportunity to see in the actual practice, as start and stop processes of a Wind Turbine should be
minimized.

We propose the use of a kind of Operator Training System specialized for each type and model of
Wind Turbine, where the trainees may observe how these big plants behave under a number of
circumstances, including when some of the elements do not have the correct behavior, but
without neither personal nor installation risks.

What is more, this type of training system may be also used to spread out relevant knowledge
gathered at any place inside the Wind Farm owner organization, reducing the cost and delay time
involved in this action. For example, information about predictive maintenance.
Data logged at 1 sample per second. 30 days period.
3 minutes zoom view of data on former figure.
Combined view of a DFIG simulator.
Drive train conditions for variable speed Wind Turbines.

DFIG on subsynchronous mode.


DFIG in supersynchronous mode.
Evolution of mechanical and electrical power under two different wind gusts.
Wind Farm SCADA

Potrebbero piacerti anche