An Understanding of Why Dual-Fuel Engines (Dfes) Are and Will Be Purchased, and How They

Introduction

An understanding of why Dual-Fuel Engines (DFEs) are and will be purchased, and how they will be used, is vital in developing markets for the technology, and predicting the impact this technology will have on carbon abatement. Predicting consumer behaviour in the DFE market is challenging as the technology is very distinct from the incumbent, and uptake to date is limited to a very small group of early adopters. As the current study relates to the impact of large scale DFE adoption, care is required when using data (which tend to be biased towards early adopters) and extrapolating to the mass market, which has very different requirements.

Much of the available data on DFE purchase behaviour and utilisation, comes from a number of trials undertaken in California, set up as a response to the Air Resources Board Zero Emission Mandate.2 Although ultimately rescinded, at the time of its introduction the ZEM posed a significant challenge to car manufacturers, prompting research into new automotive technologies, and how the buying market responds to these technologies.

Given concerns that these data sources may be biased towards US behaviour, a UK centric survey has also been developed as part of the current project. The survey focused on those who have used DFEs in the UK, and those who have demonstrated significant interest in purchasing DFEs. While not designed to provide statistically robust results (sample groups in the UK are too small) the survey represents the first publicly available systematic analysis of UK perceptions of DFEs both from the perspective of DFE owners and those who have considered DFEs.

Dual-Fuel Engines have a number of primary attributes which distinguish them from the incumbent technology. For a prospective buyer, the most important of these are higher capital cost, lower range, lower running cost, “green motoring”, make/model/brand, concerns over recharging infrastructure, and the impact of innovative technology (with regard to maintenance and resale value).

The relative importance of these attributes changes between prospective users. For example, the mass market sees capital cost as the primary factor; concerns over DFE range are also rated highly, but see “green motoring” benefits as irrelevant. In contrast, early adopters place a higher value on green motoring, will pay to support new technology, will be much less sensitive to higher capital costs and are more likely to account for lower running costs when deciding on a purchase.

Identifying the most relevant attributes for DFE adopters is important for designing effective incentives to promote uptake. When considering a purchase, capital cost is always the primary factor for the mass market. It is commonly stated that the expected lower running costs for DFEs could be used to offset a higher capital cost. However, the potential for lower running costs is largely ignored by the mass market4 i.e. life cycle costing (or Net Present Value) is rarely considered at the point of purchase. Policy makers should not overestimate the effect of an incentive policy which makes the “investment” in an EV net NPV neutral. If there is a requirement to offset higher capital costs with lower running costs, this would have to be undertaken by a third party. This gives support to the model of vehicle leasing, or pay per mile.6

The sensitivity to capital costs means that any strategy for significant uptake of DFEs (i.e. over and above the relatively small early adopter market) will need to prioritise this issue. Even a significant intervention such as a grant to cover the additional capital cost of an EV (relative to the incumbent) would not of itself be enough to ensure significant uptake. There would remain the significant disutility associated with limited range of DFEs.

Research Focus and Questionnaire Design

The pressure to switch to alternative transportation fuels exists for two primary reasons: environmental concerns and avoiding dependence on foreign

supplies of fuel. Alternative fuels exist and these sources can be exploited for use in transportation. The government has provided regulations to encourage the use of alternative fuels. Finally, the literature shows that while incentives may exist for the use of alternative fuels, the use of such fuels may be limited by such economic factors as cost of conversion, reliability, convenience, cost of operation and a resistance to change. While these factors are discussed in the literature, a significant research void remains; what are the perceptions of firms that consume and use a significant portion of transportation fuel? This paper addresses this research void by explicitly examining the perceptions held by trucking firms regarding the use of alternative fuels.

Methodology

Brace (2004) has emphasised the importance of question encoding to the success of the communication process, particularly in market research, which has to be able to successfully tune in to the language of respondents that are diverse in terms of gender and age, as well as level of education, occupation and income. Therefore, the research reported below focuses on best practice as regards question length, question wording and question order, in order to avoid negative impact on sample quality due to nonresponse – which has been shown to increase over time (deLeeuw & deHeer 2002) – or on data accuracy due to respondents’ misinterpretation of or deliberate lying in answer to questions. It should be kept in mind that good practice in terms of these issues is of particular importance in international research as it assists in reducing the impact of difference in culture and language on survey results (Brislin 1986; Smith 2003)

Question length

Results of a multi-level analysis undertaken by Holbrook et al. (2006) confirm the general advice to keep questions or statements as short as possible (Foddy 1993; Dillmann 2000; Fink 2003) in order to increase respondents’ comprehension. For the English language, Brislin (1986) specifies a maximum number of 16 words, while Oppenheim (1992) recommends 20 words per sentence whereby questions can consist of more than one sentence. In addition, Blair et al. (1977) and Andrews (1984) report increased data quality if questions or groups of questions concerning the same topic are preceded by a medium-length introduction (30 words, Blair et al. 1977; 16 to 64 words, Andrews 1984). According to evidence reported by Oksenberg and Cannell (1977, p. 342) and Jabine (1987), somewhat longer questions lead to more accurate reporting as they may convey the idea that the task is important and deserves serious effort.

Various authors (Brislin 1986; Dillman 2000; Dörnyei 2003) argue to keep the grammatical complexities to a minimum. Thus, questions should employ the active rather than the passive voice, repeat nouns instead of using pronouns and avoid possessive forms. In this way, cognitive demands on respondents are minimised and mental capacity is freed up in order to think about a response.

Another means of reducing the cognitive load on respondents stems from using specific rather than general terms (Brislin 1986; Dillmann 2000; Martin 2002; White et al. 2005), breaking down more complex questions into simpler ones (Jobe & Mingay 1989), providing behavioural illustrations of certain concepts (e.g. that a ‘chronic’ health condition means seeing a doctor two or three times for the same problem in the last 12 months; Fowler 2004) and avoiding words that indicate vagueness, such as ‘probably’, ‘maybe’ or ‘perhaps’ (Brislin 1986; Dillmann 2000). Results of multilevel analyses reported by Holbrook et al. (2006) have shown that the level of abstraction and the request for qualified judgements

increase comprehension difficulties whereby these difficulties are not dependent on age or education as is frequently assumed. Similarly, Belson (1981) and Foddy (1993) advise against the use of hypothetical questions concerning respondents’ future behaviours. Instead, it is recommended to use vignettes or alternative scenarios when seeking reactions to issues that are outside the realm of the past or present. Many studies, show that the invalidity of responses due to cognitive overload increases where recall of events is involved that have occurred not in the immediate past (i.e. more than a week/month ago) whereby the invalidity of responses depends on the importance of the event (e.g. visit to the GP vs hospitalisation; minor vs major house repairs).

While much obviously depends on the aims of the research, the target population and the special context in which a questionnaire is developed, a number of general recommendations emerge from the above review of research into questionnaire design:

• Questions should be constructed to be as clear, simple, specific and

relevant for the study’s research aims as possible.

• Questions should focus on current attitudes and very recent behaviour.

• More general questions should precede more specific questions.

• Vague quantifiers such as ‘frequently’, ‘usually’ and ‘regularly’ should

be avoided. Instead, carefully pre-tested response options should

specify the number of times per appropriate period (e.g. day, week,

month, year) of an event or behaviour.

• A desirable Likert-type response scale length ranges from five to eight

response options.

• The inclusion of a middle option increases the validity and reliability

of a response scale slightly.

• The numerical scale should be unipolar with matching verbal labels as

anchors at both ends of the scale.

• ‘Extremely’ and ‘not at all’ can serve as most effective verbal

intensifiers.

• All numeric labels should be shown to respondents.

• Numeric and verbal anchors (=endpoints) should be mentioned

explicitly.

• ‘Disagree’ options should have lower numeric values attached to them

than ‘agree’ options.

• A ‘don’t know’ option should be recorded if volunteered, whereby

interview instructions should be such that interviewers are not to

encourage respondents to choose a substantive response option if they

hesitate.

• Demographic questions should be put at the end of the questionnaire.

Of course, adherence to these recommendations for questionnaire design

will only serve to go some way in the development of a questionnaire that

is of high quality. The next step in the questionnaire design process will

be the cognitive (e.g. Jobe & Mingay 1989; Drennan 2003; Willis 2005)

and quantitative piloting (e.g. Presser & Blair 1994; DeVellis 2003; Litwin

2003) of the questionnaire in order to allow for an evaluation in terms

of its acceptance and understanding by members of the intended target

population, and an analysis of the psychometric properties (e.g. Andrich

1978; Wright & Masters 1982; Nunnally & Bernstein 1994; von der

Linden & Hambleton 1997) of its constituent questions and scales.

To gather information on the perception of alternative fuel usage at a national level it is suggested that a mailing list be obtained from The National Directory of Addresses and Telephone Numbers. This list consists of 564 firms listed under the heading of Haulage and Courier Services. A self-response mail questionnaire should be developed, pre-tested and modified. In its final form, it will be sent to the attention of the Maintenance Manager for each company listed. The questionnaire will be designed to determine the types of alternative fuels that will be used both industry-wide and by their firms in the year 2000, beliefs about enforcement of mandated changes, the importance of alternative criteria used in making a decision to convert to alternative fuels, the type of engine preferred and the degree of interest in converting to natural gas as an alternative fuel.

To protect the respondents’ confidentiality, questions pertaining to the demographic characteristics of participating firms should be kept to a minimum.

Questionnaire Structure

The first question aims at revealing the haulage company type in order to segment marketing effort at the largest single group. The groups identified are; common carriers, followed by contract carriers, private carriers and exempt carriers. Within each group we then poll the type of vehicles used in the following categories; light duty gas vehicles, light duty diesel vehicles, medium to heavy duty vehicles and medium/heavy duty diesel vehicles. This will aid further segmentation by predominant vehicle type use.

Projected alternative fuel usage. Respondents are asked which alterative fuel technology they believe will most likely to be used in the industry by the year 2010. The responses available are as follows: Clean Diesel/Reformulated Gas, Compressed Natural Gas, Gas, Liquefied Natural Gas (LNG), Other, Methanol, Electric. In a similar question, survey participants are then asked which alternative fuel technology they believe ia most likely to be used in their firm by the year 2000. The same response options are provided. Researchers (Richards, 1993; Birch, 1994; and the EIA Report, 1994) believe natural gas is the most viable fuel for the future. Based upon this assertion, respondents are then asked to rate their interest in using natural gas under the following hypothetical conditions:

1)  Cost of natural gas is £30 less per gallon than diesel or petrol

2)  Conversion costs to a Dual-Fuel Engine are approximately:

a.  Light duty £2500

b.  Class 3-5 £4000

c.  Class 6-7 £7500

d.  Class 8 £12,000

3)  Refueling facilities are provided by a local utility company

4)  Alternative fuels are not required by law.

It should be noted that these conditions represent the ‘best case scenario’ and while each condition may be found in a specific geographically niched market, they most likely do not represent what one would expect to find on a national basis. The purpose of providing the theoretical scenario is to determine whether an economic incentive

(i.e. £0.30 savings per gallon) would result in a higher level of natural gas preference.

Expected legislation enforcement/changes. Respondents are asked to indicate their

perceptions of the enforcement of the changes regarding alternative fuel use mandated by the 1990 Clean Air Act and the 1992 Energy Policy Act. The responses were: strongly enforced, loosely enforced, not enforced, and changes will be non-existent.

Key decision drivers. To determine the key decision criteria affecting conversion,

respondents are asked to rate the importance of the following items: conversion costs, availability of fueling facilities, fuel cost per mile, down time, maintenance cost per mile and refuelling time. Responses to these questions will indicate each of the factors which are viewed as critical variables in the DFE conversion decision process. Using the same criteria, respondents are then asked to estimate the payback period for converting their entire fleet to alternative fuel use based upon their perceived cost of conversion. The results will indicate whether or not there is a consistent perception of the time required for payback