Back to Library

GCI TECH NOTES©


Volume 17, Number 1           A Gossman Consulting, Inc. Publication       February, 2012


Commercial and Industrial Solid Waste Incineration (CISWI)
Cement Kiln MACT Regulations
USEPA 40 CFR Parts 60 and 241 Proposed Rules
Published December 23rd 2011

by

David Gossman

Portions adapted from comments prepared in cooperation with Universal Cement, LLC.

Introduction

On September 9, 2010, EPA published in the Federal Register a new final set of Portland Cement MACT limitations for new and existing cement plants.  On March 21, 2011, EPA published in the Federal Register a final rule which establishes MACT emission limits for cement kilns that burn solid waste.  In May of last year, GCI published a comparison of these two sets of limits and critiqued the differences.

On December 23, 2011, EPA proposed changes to the March 21 rule.  Table 1 provides an updated comparison of the emission limits for new cement kilns.  Comments on the rule itself then follow.

Table 1 - Limits from 2010/2011 Portland Cement MACT Regulations

 

PC MACT

PC Solid Waste MACT

December 2011 Proposed

Pollutant

Normal and Malfunction

Start-up /Shutdown

Normal Operations

Normal Operations

Cd

N/A

N/A

0.00048 mg/dscm @7% O2

0.00082 mg/dscm @7% O2

CO

N/A

N/A

90 ppm @7% O2

90 (straight kiln)
or 320 (PH/PC)
ppm @7% O2

PCDD/ PCDF

0.2 TEQ ng/dscm @7% O2

0.2 TEQ ng/dscm (no O2 correction)

0.0030 TEQ ng/dscm @7% O2
or 0.090
ng/dscm @7% O2

0.0036 TEQ ng/dscm @7% O2
 
or 0.075
ng/dscm @7% O2

Pb

N/A

N/A

0.0026 mg/dscm @7% O2

0.0043 mg/dscm @7% O2

Hg

21 lb/
megaton clinker

0.004 mg/dscm (no O2 correction)

0.0062 mg/dscm @7% O2

0.0037 mg/dscm @7% O2

NOx

1.5 lb/ton clinker

200 ppm @7% O2

200 ppm @7% O2

Particulate

0.04 lb/
ton clinker

0.004 gr/dscf

2.5 mg/dscm
@7% O2

8.9 mg/dscm
@7% O2

Subcategorization

The limits proposed for CISWI cement kilns are now very close to those established in the Portland Cement MACT rule.  This is not surprising given that cement kiln emissions are related to raw materials, not fuels.  Why does EPA see the need for further subcategorization of cement kilns?  The additional CISWI subcategory increases complexity and uncertainty while achieving no improvement in environmental stewardship.  A better approach would be to return cement kilns that use solid waste fuels to regulation under Portland Cement MACT.  EPA has spent tax dollars and limited resources establishing CISWI limits for cement kilns already covered under PCMACT.  Perhaps, rather than continuing this duplicative effort, EPA could simply acknowledge the limits are nearly the same and save itself the trouble of defending a second rule making.

Startup and Shutdown

Section I(C)(5) of the preamble of the proposed rule (on page 80461 FR) suggests that startup and shutdown periods are limited to 4 hours and 1 hour respectively for at least one subcategory.  This limitation may be erroneously considered appropriate for all CISWI combustion systems including cement kilns.  While the definitions of startup and shutdown in this subpart may be appropriate for incinerators or other devices, they are not appropriate for cement kilns.  Depending upon circumstances, a cement kiln startup or shutdown could be measured in days. 

Startup procedures on a kiln system depend upon the type of system and the situation (e.g. cold startup or hot startup, etc.).  In general, the system is in startup until an adequate level of production is reached that is capable of sustaining the physical, chemical and thermal process needed to make clinker.  EPA should consider startup to be the time from when fuel is initially fired until stack oxygen levels drop below 15%.

Shutdown procedures are likewise system specific, but shutdown operation would generally be considered the time from the last fuel entering the system until the ID fan is stopped.

Dioxin Quantitation Limit

Section I(C)(9) of the preamble of the proposed rule (on page 80463 FR) indicates that the dioxin limit was created using three times the laboratory detection limit as the quantitation limit for measurements.  EPA specifically invited comments on this approach in comparison with an alternative provided by a petitioner.  The EPA approach is overly simplistic and does not take into account all of the factors that are needed to properly determine a quantitation limit for dioxins.  Specifically, this approach does not account for differences in the quantitation limits for individual dioxin congeners.  It does not take into account factors that impact both the detection limit and quantitation limit for actual field samples that are subject to interferences and matrix effects, nor does this approach take into account the actual stack sampling and subsequent sample handling impacts on detection and quantitation limits in the emissions.  In order for EPA to set a scientifically defensible MACT limit on dioxins, all of these factors must be taken into account. 

In a peer reviewed paper by Rigo and Chandler, duplicate sampling trains were used to calculate the quantitation limit.  Rigo and Chandler found in this study, published by the Journal of Air and Waste Management[1], that the Reference Method Quantitation Limit for PCDD/PCDF TEQ in a municipal waste combustor may be as high as 0.51 ng/dscm.  This value is very similar to the petitioner’s suggested limit of 0.3 ng/dscm TEQ that EPA references on page 80463 FR of the preamble.

EPA should use such a detection limit value statistically derived from a large number of actual stack tests on cement kilns which takes into account the sampling and matrix effects on the quantitation limit, rather than the seemingly arbitrary approach suggested by EPA.

Dioxin Limit

Table 7 to Subpart CCCC of Part 60 and Table 8 to Subpart DDDD of Part 60 present the same limit for existing and new CISWI cement plants for dioxins on a TEQ basis but establish a radically different limit on a total dioxin basis.  This result is counter-intuitive, and may indicate a fundamental problem in the underlying calculations and analysis performed to establish these limits.  EPA should reexamine the analysis of the data for dioxins (both TEQ and total) and explain the discrepancy or clarify the limits.

60.2105 Limits

60.2105 sets a limit on CO emissions but not THC emissions.  EPA has more than 20 years of data on cement kilns that demonstrate that CO is not a good indicator of combustion conditions in cement kilns and cannot be used as a surrogate for organic HAPs.  In fact neither CO nor THC is an indicator of combustion conditions in the kiln but are impacted directly by the organics in the raw materials.  Both the Hazardous Waste Combustor MACT rule and the Portland Cement MACT rule have acknowledged this fact by setting limits on THC instead of CO, since THC is a clear surrogate for organic HAPs.  In order to be consistent in its approach to setting MACT limits on cement kilns (regardless of the MACT category), EPA should set a limit on THC rather than CO, just has been done in the Portland Cement MACT rule.

60.2105 establishes limits on lead and cadmium emissions.  In a system with relatively high exit gas temperatures (such as many incinerator systems), these metals may be volatile enough to escape from the system.  In a cement kiln, however, these metals condense in the lower temperature regions of the system and are emitted in the form of particulate.

It is likely that EPA has sufficient data, gathered in the context of the Portland cement MACT and hazardous waste combustor MACT rules, to demonstrate that emissions of these metals are correlated with particulate emissions and are independent of the type of fuel being used.  For this reason, the same logic used for nonvolatile metals in this rule (and used in promulgating the PC MACT rule) that allows the use of particulate as a surrogate for metal HAPs (other than mercury) can be applied to lead and cadmium for cement kilns regardless of the type of fuel used

60.2105 sets limits on HCl.  It is not clear that EPA has authority to set limits on HCl for kilns that are area sources.

Control Equipment

60.2110 establishes operating limits on specific emissions control equipment.  Such limits are unnecessary in cases where CEMs are used to monitor and maintain compliance. 

In a kiln system, there are multiple factors in the operation of the kiln system and air pollution control devices that impact emissions of one or more pollutants.  Controls that restrict parameters on an emission control device may unnecessarily increase emissions of another pollutant without any corresponding benefit.  For example, there are multiple ways to control NOx emissions from a cement kiln system.  Low NOx burners, staged combustion, raw material mix design and details about how an SNCR system are operated all can contribute to changes in NOx emission levels.  The proposed regulation would set operating limits on the SNCR system that would not encourage the use of alternative control strategies that could reduce the dependence on the SNCR system and thereby reduce emissions of both greenhouse gasses and fine particulate.  Similar situations arise for both SO2 and CO.  EPA should eliminate requirements to set process operating limits for emissions that are continuously monitored with a CEM, since such limits reduce flexibility without a corresponding benefit.

60.2115 establishes a complex and probably lengthy petition process for using an emissions control technology other than those listed.  This approach increases barriers to innovation and decreases incentives to use innovative and advanced technologies to reduce emissions.  Especially for those systems used to control emissions of pollutants for which a CEM is used to determine compliance, this petition process should not be required, as it adds to the regulatory burden without a corresponding benefit.

Compliance Demonstration for Mercury

60.2145(j)(3) indicates that the demonstration of initial compliance with the mercury limit be performed while the raw mill is on and while it is off.  In Table 7 to Subpart CCCC of Part 60 the mercury limit is stated as a 30 day hourly rolling average.  It is not clear that these requirements are congruent. 

The 30 day hourly rolling average limit for mercury is a health based limit that also takes into account the natural fluctuations of mercury emissions when the raw mill is shut down.  Applying this limit to a shorter time period while the raw mill is down is not appropriate and is not supported by the data used to set the limit.  EPA should make clear that the limit is not to be imposed on a specific mode of operation, but rather on the 30 day period.

Exemption for Ingredients

60.2265 exempts cement kilns that feed non-hazardous secondary “ingredients exclusively into the cold end of the kiln.”  In addition, EPA has sent a letter to the docket dated April 25, 2011 that further explains its understanding regarding the use these materials in both long kilns and preheater kiln systems.  The letter and the language in the proposed regulation do not adequately clarify the exemption.

There are many factors that determine the appropriate place for introducing materials into the kiln system.  This is especially true for those materials that have already been calcined.  These factors include particle size, emissions profile, calcination characteristics, required conveying equipment, impact on material flows, impact on gas flows, impact on temperature profile, etc.  To optimize its operation and minimize emissions (especially CO2), a facility may introduce a raw material ingredient at a variety of locations along the preheater / precalciner system or into the kiln itself.

In order to avoid confusion among the industry, permit writers, states and enforcement personnel, EPA should clearly state that kilns using non-hazardous secondary materials in the process as ingredients are exempt, regardless of the feed location.



[1] Rigo, H. G. and A. J. Chandler. 1999. “Quantitation Limits for Reference Methods 23, 26, and 29,” J. Air & Waste Manage. Assoc. 49:399-410.