GE P&W FuelSolv - Slag Control Treatment Program at a Southeastern Utility - Manual (Page 4)

Page 4 Technical Paper
3. Determining the optimum product feed rates for
the CAPP/NAPP blends while monitoring (with an
infrared (IR) camera) real-time slagging phe-
nomena along with boiler parameters such as
load, pressure drop, and exit temperature. Cus-
tomer’s fuel blends included 84 percent CAPP/16
percent NAPP, 77 percent CAPP/33 percent
NAPP, and 50 percent CAPP/50 percent NAPP as
the “highest stress test.” (the untreated “base-
line” NAPP blend slag indices were not available,
as they presented unacceptably high operation-
al risks to the customer.)
4. Demonstrating that the use of magnesium
and/or metal oxide products has no adverse ef-
fects on boiler operations- i.e., it does not exac-
erbate slagging or fouling or emissions.
Obtained Flue Gas Desulfurization (FGD)
wastewater grab samples and ash pond sam-
ples, analyzed for copper and other compo-
nents.
Data was recorded during the “baseline” (100%
CAPP coal) and chemical treatment trial periods, as-
suming equipment parameters such as tube clean-
liness, sootblower availability, and thermocouple
calibrations. Note that during the chemical trial the
unit was not derated overnight for deslagging.
Overnight load shedding allows the slag to contract
in the cooler flue gas, and this uneven contraction in
the matrix causes cracks and gravity-assisted re-
moval of the accumulated slag. This beneficial pro-
cedure was not conducted during most of the
chemical trial (Figure 3).
Here is the key chronology of the trial
 November 15- the chemical trial began as
16 percent NAPP coal was dosed at 3.0
lbs/ton magnesium-based product on the
coal belt en route to bunkering silos #5 and
#6.
 November 24-29- the trial was suspended
over Thanksgiving
 December 1 and 2- duct testing occurred.
The flue gas sampling successfully obtained
a baseline for the CAPP/NAPP blend while
the treatment consisted only of magnesium-
based product. The magnesium-based
product dosage was decreased as the metal
oxide product was introduced.
 Mid-December- following a number of
weather-related
outages, inclement
weather compelled the suspension of the
trial.
 After a weather-related equipment out-
age, NAPP coal treatment was resumed.
Dosage was increased by 10 percent after
a “gooey slag” was seen in the furnace.
The slag subsequently returned to its
semi-solid state.
 December 20- the trial concluded.
DCS Data
The trial team collected data every 5 minutes from
the customer’s “Pi” distributed control system
(DCS), including load (MW), heat rate (MMBtu/hr),
furnace exit gas temperature (FEGT, deg F), soot-
blower, and coal mill operation including feeder
flow (kpph). For the analysis, the team removed all
data points below 700 MW so that only “full load”
data was considered and outliers could not im-
pact calculations.
The most obvious change over time was soot-
blower activity. If at least one of the 60 IR (radiant)
or IK (convective) sootblowers was active when
the data was recorded, the event was logged in Pi.
In the DCS, sootblowers were either “on” or “off”; it
was not possible to record which sootblowers
were active. The trial team decided to record the
average number of sootblowers per day to under-
stand how this rate was changing over time. The
frequency seemed to have doubled as the NAPP
was increased from 0 (baseline) to 33 percent and
apparently tripled by the time the trial achieved 50
percent NAPP. It is important for a utility to keep
this parameter in mind given operations concerns
such as steam consumption and tube wear over
time.
Furnace Exit Gas Temperature (FEGT) data is valu-
able for a slagging study since it can be a proxy
indictor of slag conditions. FEGT values should rise
as slag increases, since slag is an excellent insula-
tor and because increasing slag conditions will
push the fireball farther back into the convective
pass of the furnace, this analysis indicated that
furnace operation vis-à-vis FEGT was approxi-
mately equivalent during the baseline and chemi-
cal treatment periods. The team observed
decreasing temperature variances between east
and west side thermocouples across the boiler.