3. Pellet power. Wood pellets are shown
being unloaded at the Atikokan Generating
Station during the initial test, which occurred
in January 2008. Courtesy: Ontario Power
Generation
main steam and hot reheat temperatures of
538C. Low load steam temperature control
is facilitated by flue gas recirculation. The
Atikokan GS fires lignite coal from western
Canada, delivered by rail and received at a
rotary car dumper.
The Atikokan pulverizers are of the roll-race design, originally designated MPS 75G.
These mills have a baseline coal capacity of
40. 8 Mg/h when grinding coals with a Hard-grove Index of 50 to a bulk fineness level of
70% passing 200 mesh. All of the Atikokan
mills are equipped with the original static
classifiers. Over the course of this test program, some of the mills were retrofitted with
rotating throats. These modifications from
the original static throats were conducted as
part of a maintenance upgrade but also resulted in additional valuable observations.
Atikokan’s Dedicated
Milling Concept
A small team was organized to assess the
ability of the Atikokan boiler to handle pelletized biomass via the dedicated milling
concept. Coal pulverizers have been modified
to handle wood pellets on a commercial basis
in at least three cases: Hasselby (Sweden),
Avedore 2 (Dong Energy, Denmark), and
Amer 9 (Essent, Netherlands). The Atikokan team incorporated the observations from
these projects as well as their own internal
experience with this technique at the OPG
Nanticoke GS.
When using roll-race or ball-race pulverizers to grind biomass pellets, the utility operator must consider three critical issues:
■ Limited size reduction. Coal pulverizers
depend on fracture mechanics to grind
coals to particle sizes in the 75-micron
neighborhood. However, the fibrous nature of biomass materials does not lend
itself to this mechanism. The grinding
elements in a traditional coal mill can be
expected to reduce the biomass pellet back
4. A long climb. The wood pellets are
moved at the Atikokan facility on the tripper
belt. Courtesy: Ontario Power Generation
into its constituent dust. It is critical that
the dust used to form the biomass pellets
is of a suitable particle size distribution to
allow for stable pneumatic transport and
efficient combustion.
■ Higher primary air requirements. It is
reasonable to assume that the much larger
wood particles—in the 1 to 3 mm range—
will require higher line velocities than are
employed for pulverized coal to avoid
dropout in the burner lines. OPG has employed the Rizk correlation to determine
the saltation velocity limits for a variety
of fuel/air ratios and a range of particle
sizes.
■ Cold primary air. Biomass has been shown
to release significant quantities of volatile
matter at temperatures as low as 200C. At
Atikokan, it was decided to use cold PA
to avoid the issue of early volatile matter
release. Mill inlet temperatures are held in
the 50C to 70C range for dedicated milling of wood pellets. This has been found
to be more than adequate to perform the
limited degree of drying necessary with
processed wood pellets.
Atikokan’s First
Proof-of-Concept Test
In January 2008 the first proof-of-concept
test at Atikokan was conducted. This test
employed only a single truckload of commercial grade wood pellets. Approximately
26 metric tons of pellets were delivered to
the site in “super sacs.” A simple cutting tool
was used to empty the pellets into a reclaim
hopper, where they were processed using the
existing coal-handling system without any
issues (Figures 3 and 4).
This first test was conducted at a wood
pellet flow of 5 kg/s ( 18 Mg/h) with a cold
primary airflow of 20 kg/s. The pulverizer
differential pressure while operating with
wood was observed to be much higher than
that for lignite, and the period for stabilization was also longer. The PA header pressure
was increased from 10. 5 kPa to 11. 5 kPa to
maintain the target airflow.
5. Baptism by fire. This photo shows the
unloading of the wood pellets via the bottom
hoppers of the grain cars. After being unloaded
at the Atikokan Generating Station, the pellets
underwent wood firing as part of a test program. Courtesy: Ontario Power Generation
Given the very low sulfur content of wood,
a significant reduction in SO2 emissions was
observed, as expected. However, it is notable
that the full benefit of the lower-sulfur fuel
blend did not become apparent until the mill
fully stabilized on wood.
The final key observation from this initial
short test was the relatively long clean-out
cycle required to clear the mill of wood dust
at the conclusion of testing. This phenomenon had been previously observed during
dedicated milling trials at the Nanticoke GS.
This represents a potential safety concern
over the long term if friction within this large
recirculating bed were to generate enough
heat to pose a fire hazard.
Atikokan’s Process Optimization
The next single mill test series in March 2008
had three main objectives:
■ Complete displacement of coal on a single
burner row—such as a 20% furnace energy input level.
■ Operation without the need for natural gas
support for flame stability.
■ Assess the sensitivity of NOx emissions to
the higher burner nozzle velocities associated with the firing method.
The larger fuel demands of this test required that the wood pellets be delivered by
rail. To protect the pellets from the elements,
covered grain cars were used to deliver pellets to the site. As a result, the normal rotary
dumper could not be used (Figure 5).
This larger delivery of wood pellets allowed for a longer test at the target feed rate.
Mill #3 was operated with a throughput of
6. 8 kg/s ( 24. 5 Mg/h) for this test—equivalent
to 20% of the furnace energy input. Cold primary airflow was again maintained at a base
value of 20 kg/s. Mill differential was very
stable under these conditions.
