Gossman Consulting, Inc.
August 1992

EPA CKD Sampling Methodology

In March of this year, ICF Inc. prepared for EPA a "Cement Kiln Industry Sampling and Analysis and Quality Assurance Project Plan" (Plan). As many of you may know, samples of kiln dust and clinker have been taken at a number of cement plants around the country. GCI recently took the time to review this Plan. Our observations follow.

Sampling Methodology

Section 5, pages 4,5 and 6 details the on-site sampling methodology to be used. For each type of sample taken, the Plan calls for the use of a stainless steel trowel. Stainless steel contains iron, nickel and chromium, metals that are to be determined as part of this study. Sample contamination, particularly with weather hardened CKD and clinker is probable as the samples scratch the trowel. Equipment blank provisions provided on page 10-6 of the Plan involve a water rinse only and will not reveal the type of contamination likely during sampling. Glass and/or polypropylene scoops should be specified rather than stainless steel.

Sample Preparation Methodology

Page 9 of Section 5 indicates that only clinker samples are to be ground prior to analysis. Apparently the authors of the Plan have never seen CKD samples taken from landfills. The grinding equipment is not described nor specified so it is impossible to know whether or not contamination might occur during this process as well. If a stainless steel mill is used, excessive iron, nickel and chromium values will occur. Again the homogenization equipment blank rinsate will not pick up this type of sample contamination. The Plan should specify that the grinding equipment surfaces, that contact the sample, be constructed of silicon carbide.

Page 2 of Section 8 indicates that EPA SW-846 Method 3050 will be used to prepare samples for total metals determinations (except mercury). The following comments apply to this choice of methodology.

1) Method 3050, a simple nitric acid digestion, has not been validated for use on the full range of metals specified by the Plan.

2) Method 3050 has not been validated for use on the clinker and CKD sample matrix. In fact, studies by GCI and others in the industry clearly indicates that total metal recoveries using this method are extremely poor. This fact has appeared repeatedly in the published literature*.

3) There are two published methods which have been shown to produce excellent dissolution of these sample matrixes*. These methods are a hot plate ASTM method and a microwave method.

At first glance, it might appear that no harm would come if EPA observes lower "total metals" than are actually present in CKD and clinker. However if a comparison is made between Plan "total metals" and TCLP leachable metals, the poor recovery of "total metals" could cause erroneous conclusions. These erroneous conclusions could result in restrictions on total metal inputs to cement kilns that are unreasonably and unjustifiably low.


EPA has indicated that the report from this study will be submitted to Congress and will be used to determine whether or not cement kiln dust is a hazardous waste and how it should be regulated. Industry now faces the possibility of dealing with the effects of regulations which are based upon erroneous data.

* References available upon request.