C. van Gulijk

Restriction for the ELPI in diesel particulate measurements

A dynamic phenomenon that occurs with the measurement of a diluted diesel-soot aerosol with an electrical low-pressure impactor (ELPI) was studied. An experimental setup for measuring the performance of diesel particulate filters was used for this purpose. Impactor overloading is thought to be responsible for this dynamic phenomenon. SEM analysis indicates that the fluffy or fractal structure of the diesel-soot aerosol is responsible for rapid overloading. A fluffy bed of particles on the impactor surface filters the particle-laden gas stream as it passes over the impactor surface. This phenomenon could lead to improper use of the ELPI and thereby it poses a restriction for this application. Users in the field should be aware of this phenomenon because the transient response signal from the ELPI could easily be misinterpreted as a start-up effect whereas, in reality, the ELPI is used improperly. The use of oil-soaked sintered collectors can be a solution for this problem.

Molten Salts Are Promising Catalysts. How to Apply in Practice

A diesel soot filter with a Cs2SO4⋅V2O5 molten salt diesel soot oxidation catalyst has been developed. An engine test-bench was used to test it in diesel exhaust gas with ELPI analysis and to deposit diesel soot on filters for temperature programmed oxidation experiments. Molten salt (Cs2SO4⋅V2O5) based catalytic foam has an onset temperature for catalytic oxidation of 320°C. This is a promising temperature for continuous filter-regeneration applications. Unfortunately the liquid state of the catalyst makes it unfit for the very effective wall-flow monolith filter, and necessitates the use of a foam filter as support. The onset temperature of the catalytic foam of 320°C is still too high to justify a change from wall-flow monolith to foam, as ceramic foam is a less effective filter than the wall-flow monolith. Foams are no “absolute” filters, and should be optimized for each application.

Oil-soaked sintered impactors for the ELPI in diesel particulate measurements

Diesel soot overloads the ELPI-impactor rapidly if it is equipped with the standard flat-surface impactors. This non-ideal behaviour was studied recently (J. Aerosol Sci. 32 (2001) 1117). It was found that rapid overloading, or surface build-up, is a result of a fluffy bed of soot particles that covers the impactor surfaces and starts to filter airborne soot particles. This paper reports additional results with oil-soaked sintered impactors for the ELPI. It is demonstrated that (rapid) overloading is eliminated with oil-soaked sintered impactors. The maximum allowed mass load for the ELPI impactor is increased 50-fold.

Selection and development of a reactor for diesel particulate filtration

Abstract The diesel engine is an efficient power generator but its exhaust gas needs to be cleaned. A diesel particulate filter, in fact a multi-phase multi-purpose chemical reactor for environmental protection, can lower the emission of diesel soot particles. The aim is to develop a diesel particulate filter that is robust, dependable, energy efficient and resistant to plugging. Moreover, ultrafine particles should be trapped and no poisoning gases should be emitted. A strategic approach is used for the design of a diesel particulate filter, inspired by the method of Krishna and Sie (1994) Chemical Engineering Science 49, 4029–4065, for multiphase reactor selection. The method involves three strategic levels for reactor selection: catalyst design, heat and mass injection and dispersion, and hydrodynamic flow regime; in this paper, the emphasis is on hydrodynamics. This approach has led to the design of a novel filter type for diesel soot filtration: the turbulent precipitator with foam collector plates. In this filter the gas flow is divided over two zones with different hydrodynamic characteristics: fast gas flow in an open channel and slow gas flow in stagnant zones. The open channel enables low pressure drop and prevents plugging, the stagnant zones enable deposition of diesel soot particles and, if desired, the placement of the catalytic material. The results for two different geometries of the turbulent precipitator are presented, they indicate that 90 ppi ceramic foam collector plates perform the best and that it is possible to tune different turbulent precipitators for different diesel engines. Computational fluid dynamics can be used to optimize the turbulent precipitator because it identifies the two hydrodynamic zones in the filter.

A paired comparison approach to improve the quantification of management influences in air transportation

This paper provides an approach to integrate management factors into risk analysis. The approach focuses on paired comparison quantification to differentiate and prioritize a set of management influences to reduce human or technical failure, and to quantify the size of different management influences on risk by combining it with Bayesian belief nets (BBN). The BBN model of Causal Modeling for Air Transport Safety (CATS) has been used as a case study. This approach enables better reflection in risk analysis of management functions as specific actions which can be taken by managers to maintain safety in safety critical activities.

Underestimation of language issues in frequently used accident investigation methods A new taxonomy problem found in Dutch accident data

A wide variety of methods exist in the field of accident investigation. The challenge to find the cause for each and every accident has perpetuated a complicated and fundamental debate. In spite of the different paradigms, the many branch specific investigation methods, decades of accumulated and documented accident investigation experience, the diversity in taxonomy of causal factors, and an increasing depth of general systematic analysis tools, there are still causal factors missing out. A recent study identifies language issues as an underestimated danger. Dutch Labour Inspectorate records identify procedures and communication as an important causal factor area for major accidents. Accident investigation method taxonomies that are frequently used in The Netherlands, were investigated on the content related to language issues. Language issues are found to be either present less than proportional with observed accident rates or not present at all in frequently applied accident investigation classification systems. Hence a new taxonomy problem was found in Dutch accident data. Language issues need more attention in accident investigation methods.

Experimental Techniques for the Development of the Turbulent Precipitator as a Diesel Particulate Filter

A novel type of diesel particulate filter is introduced: the turbulent precipitator. The aim is to develop a catalytically active filter, based on Cs2SO4⋅V2O5 molten salt catalyst or cerium fuel-borne catalyst. The novel filter type is developed to circumvent obvious problems like plugging and high pressure drop. In addition to that, it should be flexible, robust and possible to tune for different diesel engines. Its main features are an open flow channel (to prevent plugging and high pressure drops) and soot collection plates (to trap diesel soot). Two filter geometries are described, one with metal collector plates and one with ceramic foam collector plates. Results show that different geometries have different capabilities, making tuning for different diesel engines possible. An engine test bench was designed to measure filter efficiencies, both by particle numbers and particle mass. The diesel soot aerosol is measured with an electrical low-pressure impactor (ELPI). These measurements are not straightforward. For evaluation purposes, the engine test bench was divided into three major components to test it for aerosol measurements: diesel setup, aerosol sampling setup, and ELPI. Each part is restricted by a maximum time on stream.