UARG COMMENTS ON PROCESS INPUTS FOR EPA'S PLANNED IPM

MODELING RUNS

This document identifies several changes necessary to the input process
conditions that will be used by EPA in IPM modeling.

EPA has not explicitly identified the source of the performance and cost
algorithms, but they appear to be derived from two recent EPA publications.
These documents are:

Performance and Cost of Mercury Emission Control Technology Applications On
Electric Utility Boilers, R.K. Srivastava et. al., September 2000 (EPA-600/R-00-
083), and

Control of Mercury Emissions from Coal-fired Electric Utility Boilers: Interim
Report, J.D. Kilgroe et. al., December 2001, EPA-600/R-01-109. Specifically,
Chapter 8 of this latter document contains this information and is entitled
"Cost Evaluation of Retrofit Mercury Controls for Coal-fired Electric Utility
Boilers".

The cost and performance algorithms contained in these reports, if used for IPM
modeling, will underestimate control technology cost and overcredit mercury
removal. Accordingly, the following changes should be made to reflect authentic
utility application of mercury control equipment:

(a)	The mercury removal schedule defined in Table 1 should be utilized. This
schedule defines mercury removal for various combinations of particulate
control, flue gas desulfurization, and SCR process equipment; each as a
function of coal type.

(b)	spray cooling should not be considered a viable technology at present,

(c)	regarding activated carbon injection (ACT) in a cold-side ESP:

a.	mercury removal efficiency should be limited to 65%, per results from
the Wisconsin Energy Pleasant Prairie (PP4) demonstration,
achievable at a carbon injection rate of 12 lbs/MACF

b.	ESP capital cost must include an allowance for upgrade for small
SCA units,

c.	operating costs must include a charge for loss of ash reuse/resale

(d)	ACT mercury removal performance should follow the guidelines in Table 2,
for various combinations of control equipment.

(e)	The capital and operating cost of a fabric filter (e.g. Toxecon) is adjusted to
account for a lower air/cloth ratio (per Gaston results), and disposal of spent


-------
sorbent as hazardous material.

(f) The capital recovery factor should be consistent with either the EPRI

Technical Assessment Guide, or DOE economic premises. The application of
a simple interest and principal recovery does not completely reflect indirect
charges of ownership.

Details of these comments are described as follows:

Mercury Removal Schedule

Table 1 summarizes a proposed mercury removal schedule for various
combinations of environmental control equipment. This table reflects data from
recent pilot-scale and in-situ probe results that suggest SCR increases the
oxidation of elemental mercury for bituminous coal, but not for subbituminous
or lignite. Accordingly, the broad 35% increase in oxidation of elemental
mercury assumed by EPA for SCR for all coals is replaced by a modest increase
in mercury removal, but only for bituminous coals.

Spray Cooling

A key assumption adopted by EPA is the ability to deploy water injection for
spray cooling of flue gas to within 40°F of saturation. EPA cites minimal capital
cost for this concept ($2-4/kW), but more importantly EPA does not recognize
that spray cooling may be impractical due to corrosion induced from wetted
surfaces, or the deposition of solids that interfere with process operation. More
importantly, EPA assumes the lower process temperature (to < 300°F in some
cases) allows achieving high mercury removal (~90%) at low carbon injection
rates. Further, PP4 data showed that cooling of ESP inlet gas did not increase
mercury removal. This observation is significant as spray cooling, according to
EPA analysis can be broadly applied for 80-90% removal, at a cost of less than 1
mill/kWh.

Given the uncertainty regarding the feasibility of spray cooling, this option
should not be included as a control option in IPM modeling runs.

Activated Carbon Injection (ACI) with Cold-Side ESP

Mercury Removal Efficiency. EPA has developed mercury removal correlations
that describe the percent removed as a function of carbon injection rate and flue
gas temperature. These correlations were derived based on various pilot plant
studies and are adjusted for the inherent mercury removal documented in ICR
testing. EPA's correlations project ACI with an ESP can deliver mercury removal
up to 90% (at carbon injection rates of > 20 lbs/MACF).


-------
Mercury removal efficiency for this case is based on results from long-term data
from PP4 testing. The PP4 demonstration represents favorable process
conditions for mercury removal - a large size ESP and extended ductwork
residence time, and extremely low inherent ash carbon content. At these
extremely favorable process conditions, a maximum of 65% mercury removal
was observed, for a carbon injection rate of approximately 12 lbs/MACF. These
results are believed to represent the best-case mercury removal efficiency and
perhaps carbon utilization.

For these reasons, mercury removal from a cold-side ESP is capped at 65%, and
requires a carbon injection rate of 12 lbs/MACF.

Capital Cost for ESP. Capital costs assigned to ACI systems are assumed to
depend on the ESP specific collecting area (SCA). This assumption is based on
the observation of increased opacity with higher ash carbon content, under upset
firing conditions, and the premise that additional collection surface will mitigate
increases in opacity. For application of ACI to ESPs with smaller SCA, an
additional 1-2 fields should be added to provide additional collection surface that
can mitigate the impact of injected carbon. A capital cost algorithm is presented
(see Cost Summary, item 2a) to reflect this additional capital cost. If the IPM
model cannot reflect this degree of design detail at the unit level, then as a
surrogate for ESP SCA the unit startup dates can be utilized (See Cost Summary,
Item 2b).

Ash Revenue/Disposal Costs. The compromised ability to reuse/resell ash
should be accounted for.

The impact of higher carbon content on ash reuse and disposal options is not
clear. As indicated by the Wisconsin Energy PP4 demonstration, the ability to
reuse/resell ash can be compromised, requiring disposal by the operator.
Approximately 30% of all fly ash collected in the U.S. is sold for reuse, thus only
this fraction of operators could be affected.

The cost impact can be nationally assessed by applying a $12/ton net charge to
reflect both lost revenue and disposal costs to all units. This cost reflects 30% of
the approximate $40/ton charge cited for several units in Wisconsin for this
impact.

Maximum CO HP AC/Pulse-Jet FF Mercury Removal

Capital Cost. Commercial scale tests at Gaston with long-term data (in this case
5 days) suggest 78% mercury removal is the maximum that can be achieved with


-------
a retrofit pulse-jet fabric filter design (Toxecon). Removal of mercury at
approximately 90% was observed, but at higher carbon injection rates that could
not be sustained without incurring unacceptable flue gas pressure drop.

Mercury removals of 90% can only be achieved with a baghouse that features a
lower air/cloth ratio. Accordingly, for the IPM modeling runs, an increase in
baghouse size and cost by 20% of the "reference" case is necessary. The cost
algorithms in Table 3 reflect this necessity.

Fate Of Solid Byproducts/Waste. The spent reagent from "Toxecon" and similar
devices will feature high carbon and mercury content. It is possible this material
will be required to be treated as a hazardous waste. Accordingly, a hazardous
waste disposal charge ($1200/ton) should be assessed.

Capital Recovery Factor

The use of a capital recovery factor that reflects both interest/ principal recovery,
as well as indirect charges should be included. Ignoring the latter lowers the
fixed cost for capital recovery by approximately 25%. Both the EPRI Technical
Assessment Guide and DOE economic premises (as well as published studies by
ICF for commercial clients) utilize this methodology. Also, the operating lifetime
for this class of retrofit equipment should not exceed 20 years.


-------