Environmental Engineering Reference
In-Depth Information
adoption of appropriate measure to mitigate it would allow the combustor to run at
leaner equivalence ratio without loss of
flame. Several studies have been reported in
the past where the research community has exploited the quanti
fl
ame
behavior preceding blowout for the development of lean blowout precursors, used
for early sensing the proximity of combustor to imminent LBO. Accurate theo-
retical prediction of LBO is a major challenge due to vast complexity of the subject.
Many fundamental mechanisms of
able
fl
flame ignition, stabilization, and propagation are
still undiscovered despite the wide research conducted in the past. Phenomeno-
logical detection of LBO is found to be more promising technique in the present
time. Lefebvre (
1999
) explored the LBO detection technique based on correlation
function which was found to be very informative from design consideration of
combustor but may not be suitable in real-time prediction of LBO.
The very
fl
first among LBO-predicting strategies relied on monitoring pressure in
the combustion chamber and involved the rate of pressure drop and compressor
rotating speed (Barnum and Bell
1993
; Lucenko et al.
1996
) or monitoring the rate
of pressure
fluctuation (Snyder and Rosfjord
1998
). The major limitation of those
techniques is speed detection and complexity of algorithm.
A non-intrusive diagnostic system is, however, preferred to avoid the harsh
condition inside the combustor, and this has motivated the use of optical (Schefer
et al.
2002
; Muruganandam et al.
2003
) and acoustic (Nair et al.
2004
; Prakash et al.
2005
; Shashvat et al.
2005
) sensor-based techniques for LBO prediction. A com-
parison between the two in the same experimental setup (Muruganandam et al.
2002
; Prakash et al.
2005
) portrays optical methods as the superior choice on
account of faster response time.
Muruganandam et al. (
2005
) analyzed OH chemiluminescence to identify LBO
precursors in the
fl
flame of a premixed, swirl-stabilized, methane-fueled idealized
combustor. Thresholds were set as a percentage of the local time-averaged intensity
level for the particular
fl
fl
flame condition. The crossing of this threshold due to
fl
An increase in the frequency of
these events marked proximity to LBO conditions. Moreover, the authors imple-
ment this threshold-based sensing technique in active control system for mitigating
the LBO. They used redirecting main fuel to the pilot line as control action, which
enhances the equivalence ratio and helps to stabilize the
fluctuations was identi
ed as a
“
precursor event.
”
flame. In a parallel study
on the same swirl-stabilized dump combustor, Shashvat et al. (
2005
) developed the
active control system based on acoustic signal for LBO mitigation, thus allowing
for lean operation at lower temperature and lower emission. The authors observe
increased acoustic power
fl
fluctuation of acoustic signature
(<200 Hz) which is the result of localized extinction and reignition events. The
authors used
in low-frequency
fl
filtration of acoustic signal to produce the intermittent events, and the
events crossing the set threshold generate the control action which is again in the
form of diverting the main fuel to pilot line, which enhances the stabilization of
fl
flame in the combustor. Later, Nair and Lieuwen (
2005
) extended the development
of fast diagnostic technique to monitor the proximity of combustor to blowout using
acoustic signature from
fl
flames on three different combustors with different
fl
ame-
holding mechanisms such as pilot, swirl, and bluff-body-stabilized
fl
ames. They
Search WWH ::
Custom Search