Small is Beautiful: Evaluating Scale factors: A tool for Small and Medium Enterprises Karthikeyan Subramanian, John Olsen, Erik van Voorthuysen School of Mechanical and Manufacturing Engineering, The University of New South Wales. Keywords: Scalability, SME, Appropriate Technology, Low Cost, Scale of operation Author contact: [email protected], [email protected], [email protected]

SUMMARY

This research focuses at helping small and medium enterprises (SMEs) succeed in an environment increasingly dominated by large enterprises. Economies of scale, globalisation and mass production have pushed the scale of manufacturing processes and ability to compete beyond the scope of most SMEs. This paper describes an intuitive costing model for a manufacturing process or machine, where a comprehensive set of cost drivers can be assigned to every level in the assembly of the device as well as specific attributes of a component. An example of the latter might be a specific level of tolerance or performance for a specific component. The template for this model is the function block diagram (FBD). The model thus becomes a three-dimensional cube, where the FBD template is repeated for each layer, but contains a different cost scalability driver. The case study described in this paper deals with an ordinary stainless steel fermentation tank, as commonly used in the wine industry. Of interest is that 90% of wine producers in Australia operate at less than 10% of maximum industry scale. The design of appropriate technology to make this industry sector more competitive is therefore of paramount interest and therefore an ideal case study.

INTRODUCTION This research focuses at helping small and medium enterprises (SMEs) succeed in an environment increasingly dominated by large enterprises. Unlike large enterprises, SMEs are more constrained in terms of capital and resources, and are more susceptible to interest and price fluctuations. The aim of this work is to investigate whether scaling factors might enable SMEs to gain a competitive economic edge over large enterprises, in specific manufacturing areas. Humanity’s new found responsibility towards nature has culminated in the idea of sustainable living. This idea is influencing our approach to commodities on all the levels from the individual to the international. As a result traditional methods of costing are being reviewed and elements earlier considered “insignificant” earlier are now “taken more seriously”. Of late we have been seeing models that incorporate these “other” manufacturing in terms of their impact on society and on environment (Dieren, 1995; Perman, 2003; Suter, 2003). This approach has enabled the identification of problems associated with large scale manufacturing processes. The social and environmental costs associated with large scale manufacturing and distribution infrastructure are now considered substantial. Examples include regional unemployment, pollution, transportation costs and traffic congestion. The reemergence of discussion on the topic of downscaling and Schumacher’s “Small Is Beautiful” concept and how these may apply to manufacturing are all the more relevant now. Schumacher stated (Schumacher) that: “The most fateful error of our age is that the problem of production has been solved. This is the view held virtually by all experts in the field, managers; governments of the world …while disagreeing on almost everything all believe that the problem of production has been solved”. Schumacher felt that humanity was tending towards a process of conquering nature and dominating it. “In our endeavour for success, man has been going towards larger and bigger machinery than required and that this would be catastrophic if successful”. These advantages have unfortunately seen the demise of SMEs in most generic markets. Mahatma Gandhi, a prominent leader of the last century was perhaps first to raise an alarm on the trend of production calling for “production by the masses instead of mass production”. He felt that mass production would not be compatible with man’s need for creativity. He suggested that technology go back to the actual size of man: “We need methods and equipment which are cheap enough so that they are accessible to virtually everyone; suitable for small-scale application and compatible with man’s need for creativity. The poor in the 2

world cannot be helped by mass production, only by production by the masses; technology should go back to the actual size of man” - Mahatma Gandhi. (Ishii, 2003). Prior approaches and challenges The intermediate and alternate technology movement took off enthusiastically in the late 60’s, fuelled by Schumacher’s thought as a way towards the self-sufficiency of new economies. These efforts to introduce low-cost and appropriate technology to support local, regional employment and achieve self-sufficiency have not been successful. This is due to the rapid advances of manufacturing technology and the associated improvement in productivity, quality and cost. It was felt that the intermediate and alternative technology movement failed to deliver because firstly, it was led by well-intentioned thinkers rather than hard-nosed entrepreneurs designing for the market. Secondly it targeted singular components- such as manufacturing, machine design etc. rather than taking a holistic approach to the overall process. The Intermediate Technology Development Group (ITDG) organization founded by Schumacher changed its name to Practical Action, the virtual absence of which doomed its sister organizations. This research takes a pragmatic approach in realising this action by a model described later. Manufacturing trends Over recent decades, the competition for market share and economies of scale has bought about the growth of large scale production. The Fordian concept of mass production coupled with advanced process control, automated materials handling, sophisticated design, scheduling and costing systems has accelerated the proliferation of large, centralized and global manufacturing plants. These large plants are increasingly able to combine mass production with mass customization. In order to capture the maximum economic benefit in the early stages of new product introduction, strategies including just-in-time, Kaizen, single minute exchange of die (SMED), total quality management (TQM), cellular manufacturing and lean, flexible & agile manufacturing are all aimed at supporting the delivery of more product on the market at lower cost and in a shorter period of time. The risk of obsolete plant and equipment is partially addressed by strategies such as reconfigurable and modular manufacturing systems. Modern trends in manufacturing are more efficient, eliminate waste, minimise cycle time, offer flexibility in products within a manufacturers range. Scalability issues are related to a manufacturer’s ability to quickly changeover and produce say a million parts of model A, two million of model B, three million of model C, than the conventional method of producing just

3

six million of model A alone. Narasimhan (Narasimhan, 2006) has complied references for lean, flexible & agile manufacturing. All the approaches are towards high volume production. Modern Technological trends enabling SME’s There has been a dramatic decrease in the cost of equipment like computers, robots, control systems, sensor technologies and vision systems over the last few decades. Figure 1 shows that the cost of computing has declined by approximately four orders of magnitude in the last

Cost $

six decades. 1,000,000 100,000

PC Price Index

10,000 1,000 100

CPI (US)

10 1 1959

1964

1969

1974

1979

1984

1989

1994

1999

2004

Year

Figure 1 Historic Trend for the cost of computing (The figure is adapted from Stiroh, 2007) Similar technological advances in steel making made possible cost-efficient manufacturing processes that previously have been beyond the reach of SMEs (Johansson and Holappa, 2004). The effects of environmental pressures and awareness as well as the increasing cost of raw materials have forced these industries to look for efficiencies beyond “economies of scale” as appropriate technology became available. Historically, growth in industry was bought about by improvements in productivity through an increase in economies of scale; According to (Johansson & Holappa 2004) development efforts have almost always been “diverted almost entirely towards overcoming the technical obstacles limiting further size increases rather than towards innovative process alternatives”. The development of the electric arc furnace enabled the development of mini-mills in the steel industry. Furnace technology, ladle metallurgy and casting in sequence resulted in both high productivity and efficiency. It also allowed for a high proportion of scrap to be recycled 4

relying less on the primary process from ore. A similar trend is found in the paper industry where large amounts of recycled pulp can be processed in mini-mills, again reducing the dependence on primary processes. Research studies involving pilot scale laboratory trials and theoretical calculations conducted on mini-mills using recycled paper instead of virgin pulp, have shown a number of improvements. Firstly, the mini-mill uses less than half the electrical energy of current best available technology (350kwH per tonne of pulp), it is more thermally self-sufficient (a fuel efficiency of 70%), it enables the recovery of more of the pulping chemicals (85-90%) and it produces an eucalyptus pulp substitute from straw in 30-40 minutes, ready for bleaching. (Riddlestone, 2001). Have these trends complemented SME’s? Is there a systemic way of mapping technological advances and evaluating both user and social requirements while meeting the desired product performance requirements, quality standards, profitability and regulations?. Most of these concepts are well beyond the scope of SME’s whereas policies are formulated invariably in favour of large organisations. SCOPE OF THIS WORK The aim of the present research is to develop a model that will allow the scalability of a given industrial process to be analysed and calculated. This study has developed a methodology that is capable of identifying key scale and process design drivers. It incorporates dynamic trends to model how these may change over time. The outcome is a scalability index that may be applied during the conceptual design stage and subsequently updated to adapt to new opportunities. The next step in the formulation of such a methodology is the development of a cost model that can be universally applied to different types of processes. Manufacturing Cost Model In order to provide a link between the manufacturing Key Factors of Success (KFS) and process design drivers it is necessary to apply a manufacturing cost model and to demonstrate that the key cost drivers of a process are intimately linked with product and process design characteristics. In particular it is important to establish the idea value added (VA) as an economic interpretation of the complexity of a manufacturing process and therefore process design. Kalpakjian & Schmid (2001) provide an economic and technical representation of a production process. This is shown in figure 2.

5

Figure 2: The manufacturing process as a (a) economic and (b) technical system (Kalpakjian & Schmid (2001)). The total cost of product leaving the manufacturing system is equal to the sum of raw material costs entering the system, the labour costs contributed by the organisation and the cost of capital. Value added is the ratio between the costs that are contributed by the manufacturing process (labour and capital) divided by the total cost. Value added includes all costs added to the product, whether these costs are in fact “value adding” or “cost adding”. The efficiency of the use of capital resources is given by the ratio of total cost and investment in capital, which in the case of manufacturing includes process plant and equipment, buildings, services and inventory. We give this ratio the name “capital velocity” because the velocity vector reflects two important characteristics, namely speed and direction. The capital, speed is indicative of how efficiently capital is employed in the manufacturing system. The direction of the capital velocity vector reflects how well the nature and capability of the asset base is aligned with the current and future direction of the manufacturing strategy associated with the system. The study conducted by van Breukelen, et al. (van Breukelen, 2000) has established the presence of a strong relationship between capital velocity and value added for modern, automated manufacturing systems. This relationship is shown in figure 3. These manufacturing systems incorporate automation technology to achieve the drivers for the key factors of manufacturing success as opposed to a reliance on cheap manual labour. Implicit in the data shown is the assumption that a certain level of productivity has been achieved such that the VA figures are not unnecessarily inflated with unproductive costs. Figure 4 illustrates the concept of the industry cost curve for product that is predominantly generic and where little differentiation exists between suppliers. The high-volume, low-cost producers “manage” supply and price. The single shaded producer on the right (E) is only marginally profitable, whereas the producer on the far right (F) is unprofitable.

6

The significance of the relationship derived in Figure 3 is that for a given level of value added (read “process complexity” or “industry activity”) there is an optimum level of investment in capital assets. The premise of this argument is that regardless how desirable it may be to up or downscale a process, basic constraints such as the level of capital investment must be maintained (figure 5) in order to remain competitive, now and in the future.

Figure 3: Capital Velocity as a function of Value Added (adapted from van Breukelen et al, 2000)

Figure 4: The Industry Cost Curve Concept

7

Figure 5: Calculation of Value Added and Capital Velocity

8

Process Design Drivers For the purpose of establishing a basic set of process design drivers it is useful to further develop the abstract shown in Figure 2b. This is shown in Figure 6. This model is an abstract representation of a generic machine process and the factors that influence its design and performance.

Figure 6: Abstract representation of an Industrial Machine A manufacturing process consists of a number of building blocks, including process equipment (machinery), associated automation and materials handling, a suitable layout and topography (e.g. a flexible manufacturing cell), and control software implemented at different levels within the process (device level, cell level, etc.). The selection process for each of these building blocks depends on the volume of product to be manufactured and on the flexibility required. Flexibility is a process characteristic that reflects the ability of the process to absorb changes in design, quantity, product mix, etc. without undue economic penalty. This is shown graphically in Figure 7.

9

Figure 7: Manufacturing Automation as a function of Volume & Variety (adapted from Groover, 2001) The essential design drivers can be summarized as follows:  The product: its design, functionality and features translated into manufacturing steps and operations (through QFD)  The level of value added (VA)  The level of demand, and its projected growth  Supply and service criteria  The lifecycle of the product, and future design variations and trends  Benchmark cost range for similar products offered by competitors. o An alternative to the benchmark cost is the estimated level of revenue less a gross margin (say 40%) divided by the annualized volume of production.  Level of product and process technology, including automation and control.  Cost of labour vs. automation o Benchmarks cost indicators for actuators, sensors, controllers, computer hardware, software  Cost of mechanical systems o Expressed on the basis of a “generic mix” of mechanical elements, $/kg, $/unit of output (structure, bearings, gearing, valves, etc.)

10

 Cost of support services  Labour requirement (local and international) o Skills required, staffing, cost The abovementioned drivers are “semi-dynamic” in the sense that they are not expected to remain constant over time. However they are not expected to behave in an entirely stochastic way either. Rather a forecasted trend line is associated with each of these variables and this trend line may be represented mathematically. An example of such a trend is the decreasing price of computer hardware as shown in Figure 1. The design decisions include the following:  Selection of manufacturing process and technology  Required flexibility of the process  Size or scale of the mechanical system o This can be expressed in terms of physical size, production throughput, the scale versus cost of the system.  Materials handling and storage requirements  The level of automation: the degree and extent of actuation, sensing and control o Within the process o External to the process o Grade of automation (short, medium or long term)  Support services required, and their scale In addition, the design decisions are subject to a number of constraints, including:  Capital cost is constrained by the CV-VA equation  Return on investment and payback time  Automation and materials handling costs are included in the above constraint. As a rule, automation and materials handling contribute around 30-40% of total capital outlay.  The unit production cost must be within, say + 5% of benchmark cost.  Process reliability is an important consideration in terms of meeting demand and supply constraints and a suitable target could be, say 95%.  Quality, again a target will be set in line with best practice and process capability  Capacity must be greater than demand  Value added constraint (depending on the market segment, product positioning, etc.).

11

Since the aim of the study is to formulate a framework that is capable of assigning a “scalability index” to a process, each of these constraints will contribute to the index. To determine whether these technologies give an economic advantage to SMEs over large enterprises, we define and identify sets of scalability indices for various processes. These can act as indicators as to the extent to which a process or product could be economically scaled down. THE SCALABILITY INDEX MODEL Considering the challenges discussed above any meaningful approach to aid SME’s needs a holistic approach that looks at a high level vision augmented by a practical model that is 1. Comprehensive 2. Flexible 3. Adaptable to changing conditions, manufacturing, social, regulatory etc. 4. Easy to use The approach is modeled and evaluated in the following levels 1. A high level model that is holistic and considers all relevant parameters 2. The Function Block Diagram (FBD) that acts as a template for identifying and classifying complex systems down to component level. 3. Case studies on process/product having a wide range of manufacturing volumes and sizes. This will be discussed next.

12

The high level model The scalability index measures the extent to which a process can be scaled down. Some scalability indices are measurable directly while some are not. As mentioned earlier of those that can be measured, some until recently have not been considered important (Dieren, W. v.). The awareness of the environmental impact of manufacturing processes and the cost associated with them is one obvious example. To address this problem, a holistic model that evaluates scalability based on technology, micro & macro-economic criteria, environmental, social, and technology trends is developed (Fig 8).

Figure 8: Holistic Model to evaluate Scalability The model is designed to incorporate not only standard industry cost models but also socioenvironmental variables and variables related to the design of manufacturing processes

13

The model identifies a wide range of categories that will be addressed in this study. Some of the main classifications and interdependent variables are shown in Table 1 below: High-Level Relationship

Attribute

Cost Driver

KFS

Manufacturing

Operations

Labour, Productivity.

Labour - Automation Balance

Process

Capital Investment

Flexibility, Capability, Capacity

Match Quality -Capability Align Capacity Demand

Technology

Appropriate Functionality

Complexity, Training

Trend, Product - Process (QFD)

Market

Demand, Price Volume/Price Sensitivity, Growth

Productivity, Profitability, Differentiability

Environmental

Recycling, Emissions, Waste

Energy, Waste, Carbon Tax.

Efficiency

Transport

Non-value added cost

Distance, Weight, Volume, Pollution

Reduce distance, Focused market (regional)

Socio-economic

Social Costs

Unemployment

Regional employment opportunities

Table 1: Attributes between high level relationship and KFS considered in the model. Function Block Diagram (FBD) The FBD is a hierarchical, step-by-step breakdown of complex functional groups into increasingly more detailed systems, equipment items and components. This enables the identification of tasks by functional boundaries displaying in a logical and sequential relationship that that can be used as a template. It is a natural expansion of the abstract of the machinery (Figure 6) and can produce an accurate and comprehensive listing of parts of any machine categorised into its operating functions. At the lowest level the FBD lists the individual components of the Bill of Materials (BOM) with assemblies and sub-assemblies at the intermediate levels. They can be expanded to a series of levels as required ( see Figure 9).

14

Vessel FBD

Structure

Vessel

SS Sheet

Cooling jacket

Support

Lid top & bottom

Stand

Brackets

Inlet Valve

Flow System

Control

Piping

Control Mechanism

Outlet Valve

Drain Pipe

Drain Door

Visual

Flow

Heat

Power

Pressure

Flow

Cooling

Pump

Heat Exchanger

Figure 9: Function Block Diagram (FBD) Wine Fermentation Tank (Vessel) The three dimensional (3D) representation of the cost model is multi-dimensional in that:  It considers all relevant parameters, variables and attributes, e.g.: 

Environmental (e.g. impact of carbon tax)



Social (labour, employment)



Technology trend (impact of new technology)



Material (influence on choice of materials/recycled products)



Process (influence of tighter tolerances and optimum product design)



Recycling (environmental impact)

It is also dynamic in the sense that it can take into account changing trends and conditions. The above diagram lists one layer of the FBD which is attributed to a variable to be indexed. Based on the number of variables the FBD is expanded to “n” layers each of which represents a design cost driver. The FBD can be used as a template in a 3D cube (Figure 10), where each layer represents a variable or parameter that contributes to the cost of the design. The FBD structure is repeated for each layer so that the design cost can be analysed by drivers (e.g. material or labour) as well as by equipment item or component.

15

ΣM ΣL ΣCC

Assembly Scalability

Cost Driver

ΣPA

Functional Scalability

Different layers abstract to building multi dimensional FBD to evaluate Scalability ΣM = Material, ΣL= Labour, ΣCC = Capital Cost, ΣPA = Purchased Assemblies (OEM), Σ… etc, etc

Σetc.

Figure 10: Different FBD Layers contributing to the Scalability Index The model is three dimensional (3D) with the x, y & z axes indicating component, assembly and attribute values. CASE STUDY Identification of industry for Case Study A case study of small-scale wine production was chosen because:: 1. There are a large number of producers with varying capacities of production, 2. They produce similar product, 3. Process from raw material to a finished product, 4. Have a variety of process, 5. Have critical process requiring innovative solutions and machinery, 6. Is widespread and has production in various conditions, climatic, regional, remote and close to cities, 16

7. Growth in the higher and lower end of the chain both in value and quantity of production. Overview of Wine Industry in Australia  2399 producers;  90% operate <10% of max industry operating scale;  70% operate < 1% of max industry operating scale.

Tonnes crushed Less than 50 50 to 99 100 to 249 250 to 499 500 to 999 1,000 to 2,499 2,500 to 4,999 5,000 to 9,999 10,000 Unknown Total

2002 749 212 189 88 61 54 36 23 41 12

2003 866 230 199 103 57 56 36 27 43 8

2004 996 254 211 106 72 45 40 24 43 7

2005 1028 288 229 111 76 55 33 29 41 9

2006 1082 33 242 126 74 69 27 28 41 16

2007 1185 322 235 133 73 70 30 26 33 40

2008 1280 330 343 146 74 67 36 26 35 62

% 53.4 13.8 14.3 6.1 3.1 2.8 1.5 1.1 1.5 2.6

1465

1625

1798

1899

1738

2147

2399

100.0

Table 2: Number of wine producers by tonnes crushed since 2002 (Australia)

Table 2 (Annual financial benchmarking survey for Australian wine industry Vintage 2005, Report: Winemakers Federation of Australia) indicates a continuing growth in small wine manufacturers. Of all the producers nearly 90% operate on scales less than 10% of the maximum industry operating scale yet the technical barriers for small wineries are quite significant as listed below:

17

Technical Barriers for Small Wineries  Centrifuges  Filtration devices  Wastewater treatment facilities  Bottling facilities  Increased wine losses due to small inefficient batch operations  Unaffordable technical expertise in process, wine chemistry and sustainable production.  Crushing above 5 tonnes labour intensive; uneconomical to automate as automation for this volume is expensive.  Cost of equipment is not proportionately scalable. That is 1 ton press costs half that of a 10 ton press.  Transfer of wine into refrigerator to be done immediately after crushing restrained by scale factors often leads to loss of quality and unable to meet OH&S regulations. Product for evaluation: The wine fermentation tank (Figure 11), referred to as the vessel, is chosen for the pilot study. The capacity of a tank varies from a few hundred litres (as commonly used by small wineries) to several tens of thousands of litres in the case of larger wineries. A FBD model is created by establishing, in the first instance, the major functional groups and systems. Next, the subsystems, equipment items and components are identified. Certain equipment items are purchased as entire assemblies from external suppliers. Depending on the scope of the design, these items do not need to be developed any further in the FBD hierarchy. The manufacturing tolerance is not considered critical in this product however the cleanliness and hygiene are critical across the range.

18

Figure 11: Sectional view wine fermentation tank The cost has been compiled based on commercial sources from Australia, US and China. Data and analysis and discussion Table 3 shows the cost for various sizes. The scalability cost drivers are restricted at this point to material and labour as a function of size. Purchased assemblies including the refrigeration unit, pump, heat exchangers and valves are treated as “off-the-shelf” and appropriately costed. It is to be noted that each of these parts and assemblies could be subjected to their own FBD analysis which would be able to evaluate their Scalability Index.

19

Manufacturing process and cost of tank in different volumes Unit value Total cost Variable Process Unit Measure Cost 5000lts 10000lts 20000lts 5000lts 10000lts 20000lts Weight kgs 1050 1670 3150 Raw material Size SqM 180/225 10 20 25 $1,800 $3,600 $5,625 Thickness mm 3 3 4 Shearing/cutting Hours $/hour 60 1 2 4 $60 $120 $240 Rolling Hours $/hour 75 2 3 5 $150 $225 $375 Welding Hours $/hour 65 45 70 82 $2,925 $4,550 $5,330 Labour/ Forming top Total cost $ 1500 1800 2500 $1,500 $1,800 $2,500 fabrication Forming bottom Total cost $ 1500 1800 2500 $1,500 $1,800 $2,500 cost Assembly accessories Hours $/hour 50 14 14 20 $700 $700 $1,000 Cleaning Hours $/hour 40 12 12 15 $480 $480 $600 Erection Hours $/hour 60 3 3 6 $180 $180 $360 Cleaning Hours $/hour 50 4 4 4 $200 $200 $200 Total labour/fabrication cost $7,695 $10,055 $13,105 Inspection door Bought out 45 45 45 $45 $45 $45 Raking door Bought out 750 750 750 $750 $750 $750 Bought out Pump Bought out 3150 3150 4000 $3,150 $3,150 $4,000 Controller: PLC Bought out 450 450 450 $450 $450 $450 Cooling jacket Bought out 2500 3000 4200 $2,500 $3,000 $4,200 Tolerance Volume Not critical Manufacturing Size Not critical Temp specifications control Critical Total cost $16,390 $21,050 $28,175 Dimensions

1600*3300 2000*4500 2150*6300

Table 3: Material and cost comparison of manufacturing of Tank in various sizes

Cost $/liter

Cost /liter, 5000, 3.28

Cost /liter, 10000, 2.11 Cost /liter, 20000, 1.41

Size of Vessel in Litres

Figure 12: Aggregate scalability expressed in cost/litre Figure 12 shows the overall cost per litre for a stainless steel fermentation vessel.

20

An analysis of the individual costs based on the FBD model shows those items that are sensitive to scale and those that are not. This is shown in Figure 13 (as a total cost) and Figure 14 (as a cost per litre).

Cost $

6,000 Sheet Cost Welding Cost

5,000

OEM Cooling Jacket OEM Pump

4,000

3,000

Dome Pressing

2,000

1,000

Assembly Acessories OEM Racking Door Cleaning 1 PLC COntroller

0 0

5000

10000

15000

Cleaning 2 OEM Inspection Door 20000 25000

Size of vessel in litres Figure 13: Individual Scalability (Individual cost of process and components) It can be seen that Material used Stainless Sheet is scalable with a cost variation from 1800 to 5,625 Welding cost door is scalable with a variation of 2,995 to 5,330 21

Cooling jacket is less scalable with value from 3,000 to 4,200 Whereas inspection doors, valves sampling vent are not significantly scalable with cost almost constant throughout the range selected.

Cost per liter $/liter

0.70 OEM Pump Cost 0.60

0.50

OEM Cooling Jacket Welding Cost

0.40

0.30

Sheet Cost

0.20 OEM Raking Door 0.10

Dome Pressing

OEM PLC Low Cost Constant Value Items

Assembly Accessories

0.00 5000

10000

20000

Size of Vessel in Litres Figure 14: Economies of Scale (Cost per litre) Based on the model and graph the equation for some variables are interpolated as

x

Value y

Regression R2

Process : Welding Material: Sheet metal cost

y = 2205.8ln(x) + 2950.9 y = 1912.5x - 150

0.9975 0.9988

OEM: Cooling Jacket

y = 1879.9 e0.2594x

0.9714

2

1

2

1

OEM: Pump

y = 425x - 1275x + 4000

Process : Dome Pressing

y = 200x - 300x + 1600

OEM: Low cost purchased

y = 705

Total Cost

y = 12415e0.2709x

Constant 0.9981

22

Although data was collected from a wide range of tank capacities from 500 litres to 20,000 litres the comparison details were not available for all sizes. Hence the data is presented for a range between 5,000 and 20,000 litres. The model lists costs that are scalable and linear, scalable and non-linear and non-scalable (or constant) across the range. The need and ability to maintain close tolerances has a significant impact on the cost, volume and size of the product/process. Being a simple process and product the dimensional accuracy is not a significant factor in wine fermentation tank but maintenance of temperature and cleanliness is. This is reflected in the exponential nature of the cost of the cooling jacket whereas the sheet metal cost varies rather linearly. The design and manufacture of the cooling jacket and refrigeration system is another FBD analysis which is not considered here and considered as a purchased item (OEM). Other components, valves and doors, which are constant irrespective of the size of the vessel, within this range are constant and reflected in the graph. There is hardly any scope in attempting to scale these items. Conclusion This paper discusses a design tool that is, in the first instance, aimed at analysing the main functional design requirements and to establish the main cost drivers for each functional group as a function of the scale of the device. There are many cost drivers. The obvious ones of course are material, labour, depreciation and lease (capital ownership), transport and energy. Less obvious ones would include carbon emission, waste, disposal as well as various social costs.

Key design parameters that the tool should address include the level of

functionality, the scale of the device and the capability (tolerance or control levels) of the process.

The philosophy behind the model rests on key principles borrowed from

RCM/FMECA (the function and system block diagrams), value engineering (functional design analysis and simplification), life-cycle costing (identification of all through life costs to disposal), QFD (alignment of functional and technical specifications with commercial requirements) and Taguchi analysis (robust design specification ranges for scale and capability). This design tool is not meant to be used in isolation nor is it a comprehensive tool in its own right. It is a scalability analysis tool that should be used in conjunction with well-established design processes, such as the traditional Pahl and Beitz process or more recent design protocols such as VDI-2206 and 2221(Verein Deutscher, 2004). 23

References Annual financial benchmarking survey for Australian wine industry Vintage 2005; Publication date: June 2006: A joint project of Deloitte and the Winemakers’ Federation of Australia. Dieren, W. v. (1995). Taking nature into account: a report to the Club of Rome: toward a sustainable national income. New York, Copernicus. Groover, M. P. (2001). Automation, production systems, and computer-integrated manufacturing, Prentice-Hall. Ishii, K. (2003). An economics for development and peace: with a particular focus on the thought of Ernst F. Schumacher, Springer. Johansson, A. and L. Holappa (2004). "From megaplants to mini-mills - a trend in steelmaking - a prospect for papermaking." Resources, Conservation and Recycling, 40,173-183. Kalpakjian, S. and Schmid S. R. (2001). Manufacturing engineering and technology, Prentice-Hall, New Jersey. Narasimhan.R, "Disentangling leanness and agility: An empirical investigation," Journal of Operations Management, vol. 24, pp. 440-457, 2006. Perman, R. (2003). Natural resource and environmental economics. Harlow, England, Pearson Education. Riddlestone, S. (2001). Case Study: The MiniMill Concept. The Paper Industry Research Association (PIRA) Conference “Cost Effectively manufacturing Paper and Paperboard from Non-Wood Fibres and Crop Residues” Amsterdam, The Netherlands Schumacher, E. F. (1999). Small is beautiful: a study of economics as if people mattered. Point Roberts, Wash., Hartley & Marks Publishers. Sebastian K. Fixson ,Assessing Product Architecture Costing: Product Life Cycles, Allocation Rules, and Cost Models, ASME Conf. Proc. 2004, 857 (2004),DOI:10.1115/DETC2004-57458 (cite). Stiroh.K, Federal Reserve Bank of New York, CESifo Economic Studies Conference: "Productivity and Growth", June 22, 2007 Suter, K. (2003). Making the Environment Count. Contemporary Review. 282: 222227. van Breukelen, Q. H., Koolhaas C.B and Kumpe, T, (2000). Benchmarken van industriële processen: resultaten van een wereldwijd onderzoek naar de operationele prestaties van industrieën, Uitgeverij Van Gorcum. Verein Deutscher, Ingenieure, Design methodology for mechatronic systems : VDI 2206, Berlin : Beuth Verlag GmBH, 2004 Winemakers Federation of Australia, National Wine Centre, Botanic Road, Adelaide, SA 5000.

24