Technical Marketing Optimisation for Coal

Q: What is technical marketing optimisation for coal?

It is the use of optimisation to test whether a mine’s coal product definitions are well aligned with the coal it produces and the markets it sells into. BlendOpt values each candidate product set against the mining schedule, washability, CHPP and logistics constraints, sales contracts and price forecasts. This lets technical marketing build a business case for product changes that operations can actually deliver.

Q: How can changing a product specification increase profit?

In this study, replacing product 5950 with a new product 6100 required a 150 kcal/kg (2.7%) increase in minimum CV. BlendOpt found an optimised product strategy with an 8.4% increase in total profit over eight months. The analysis suggests the gain came from better alignment between product definitions and the Ash-CV-Yield relationship of the mined coal.

Q: Can fewer coal products be as profitable as more?

Not always, but in this study two new product definitions replaced all five existing export products with an eight-month profit improvement of 14.3%. This is notable because fewer unique products are usually expected to reduce the profitability that product diversity offers. A smaller product set also brings practical benefits through simpler operations and business processes.

Q: How are prices set for new coal products that are not yet traded?

For the Newcastle indexed thermal products in this study, pricing discounts had a strict linear relationship with product CV acceptance ranges, so new products were priced on the same linear function. Other quality ranges such as moisture, VM and sulphur were interpolated from the two current products most similar in energy. New products were defined by equidistant interpolation between existing products, which kept the number of candidates manageable.

Q: How reliable are the results of a technical marketing study?

The results depend on the accuracy of the financial model, and this study found they were sensitive to cost assumptions where costs approach revenue for some products. Before sharing the findings more widely, the client used an independent third party consultancy to validate them. The client also used its BlendOpt licence to extend the study with a more comprehensive cost model, alternative price forecast scenarios and additional mine schedules.

What BlendOpt optimises

What BlendOpt optimises in coal technical marketing

Each product strategy is valued against the mining schedule, washability, CHPP and logistics constraints, sales contracts and price forecasts together, so marketing and operational planning stay aligned.

01 / Specifications

Product specification changes

Test new quality acceptance ranges for export thermal products, such as a higher minimum CV, and see the effect on total profit.

02 / Consolidation

Product consolidation

Assess whether fewer products can replace the existing product set while achieving a financially similar or better outcome.

03 / Pricing

Price and quality relationships

Apply product pricing forecasts and quality-based price adjustment factors so every product definition is valued consistently.

04 / Coal quality

Coal quality and washability

Align product definitions with the energy, ash and washability of the coal in the current mining schedule.

05 / Constraints

Value chain and contract constraints

Include ROM stockpile blending, CHPP rates, stockyard blending, conveyor and railing limits and committed sales contracts in every scenario.

06 / Scenarios

Scenario and sensitivity analysis

Rerun the study under alternative cost models, price forecast scenarios and mine schedules to check that results hold.

Customer results

Technical marketing optimisation results

Results BlendOpt customers have measured in product strategy, blend value and product quality at coal operations.

+8.4% overall profit improvement from a 2.7% increase to CV in the highest volume thermal product Read the case study below → Over 30% higher value for some blends at a South African coal exporter, within contract requirements Port blending and day-of-operations scheduling → 35% → <5% coal quality specification violations across a five-year plan Coal quality consistency case study →

The whole value chain

Technical marketing in the context of the whole value chain

A product specification only earns value if the mine, the CHPP and the logistics chain can deliver it.

BlendOpt takes the mining schedule as an input and models stockpiles, washability, processing, rail and sales contracts together, so each change to the product set is valued by its effect on the whole operation. See how this works for coal blending, product portfolio optimisation, train-to-vessel blend scheduling and integrated mine planning and scheduling.

Marketing meets operationsProducts the mine can deliver.

BlendOpt model

Mineschedule & costs
ROMstockpile blending
CHPPwashability
Railconveyor & rail
Marketcontracts & prices

Optimisation objective Align marketing and operations

Each candidate product set is valued against what the mine, CHPP and logistics chain can deliver, and what the market pays for it.

Customer case study · Coal

Case study

Export thermal coal products

8.4% overall profit improvement from a small change to one product specification

A New South Wales coal mine used BlendOpt to test changes to its export thermal coal product definitions and found that two new products could replace all five existing ones.

+8.4%profit, one spec change
5 → 2export products
+14.3%profit, consolidated set

Independently validated

Three-day study

A time-boxed study built the business case; the client then had an independent third party consultancy validate the findings.

See the value unlocked →

Full case study

The challenge, solution, study results and methodology for a NSW coal mine’s product definitions, as a PDF.

PDF Download the case studyCoal technical marketing optimisation (PDF)

Highlights

8.4% overall profit improvement
Products reduced from 5 to 2
Findings validated by an independent third party
Study extended with the client’s BlendOpt licence

01

The challenge

A coal mine in New South Wales, Australia wanted to investigate the financial impact of modifications to its export product definitions, as well as the financial impact of export product consolidation.

The Technical Marketing team wanted to understand whether the existing thermal export product definitions were well aligned with the mine’s distribution of coal energy, ash and washability, given its current negotiated mining schedule, value chain constraints and various contractual commitments.

Technical Marketing also wished to investigate whether a smaller, consolidated product set could replace the existing product set while achieving a financially similar outcome for the mine.

Product consolidation was a significant area of interest, as it was expected to bring many practical benefits through operational and business process simplification.

02

The solution

The aim of this project was to undertake an accurate technical marketing study to align marketing and operational planning. The study was run as a three-day, time-boxed exercise to demonstrate how BlendOpt, configured for integrated planning, can be used to rapidly explore a specific question and build a business case for improving alignment between market conditions and value chain operations.

Experiments focused on adjustments to the quality specifications of export thermal coal products. The chosen set of product adjustments was motivated by an initial analysis of the distribution of optimised product volumes and the associated product definitions (see Study methodology).

A BlendOpt integrated planning model was developed for this client, including several model constraints found to be essential to accurately aligning marketing and operational planning.

The integrated planning model included client mine scheduling data and costs, ROM stockpile blending and constraints, crushing and CHPP processing rates and availability, coal washability data, post-CHPP stockyard blending models, conveyor and railing constraints, committed sales contracts, product quality acceptance ranges, product pricing forecasts, and quality-based price adjustment factors. A simplified illustration of the value chain is shown in Figure 1.

Figure 1: integrated value chain model. Coal from two pits is blended onto ROM stockpiles, passes through the crusher and CHPP onto processed coal stockpiles, and is blended to product specification for a power station or railed to port for committed and uncommitted sales — Figure 1. Integrated value chain model.

This integrated planning model was designed to support a range of strategic and operational business decisions and planning processes for the client, so some model features were not critical to the findings of this particular study.

03

The value unlocked

The integrated planning model used in this study accounted for detailed operational and business constraints, which enabled a high degree of financial estimation accuracy within a technical marketing study. Considerations included published coal mining schedules, dynamic CHPP and supply chain modelling, coal washability data, sales contracts, product quality targets and acceptance ranges, price forecasts, and quality-based price adjustment factors.

The following noteworthy observations resulted from this study:

A 2.7% increase to CV in the highest volume thermal product corresponded with an 8.4% overall profit improvement for the coal mine.
Two new product definitions were able to replace all five existing export products, with a remarkable eight-month profit improvement of 14.3%.

The product consolidation results are particularly noteworthy given the trade-offs typically expected between reducing the number of unique products and the theoretical profitability afforded by product diversity.

Technical detail

Study results

A number of experiments were conducted in which one or more existing products were replaced with new products having a modified quality acceptance specification and an associated change in pricing forecast (see Study methodology). Presenting all results is outside the scope of this report, and only selected results are reported in this section.

Among the product replacement experiments, the most noteworthy are the single product and two product replacement experiments reported in Table 1. For experiment EXP-7 (Single Product Replacement), product 5950 was replaced with a new product 6100. The new product specification requires a small 150 kcal/kg (2.7%) increase in minimum CV. Running BlendOpt on this new problem definition resulted in an optimised product strategy with an 8.4% increase in total profit over eight months.

Experiment EXP-7 displayed the most pronounced profit uplift; however, other single product replacement experiments also exhibited opportunities for improvement. For experiment EXP-3 (Two Product Replacement), product 5950 was replaced with a new product 6100 and product 6200 was replaced with a new product 6300, resulting in an 11.4% increase in total profit over eight months.

Table 1: performance metrics of optimised plans. Profit increase over the baseline was 8.39% for single product replacement (EXP-7), 11.65% for two product replacement (EXP-3) and 14.29% for product consolidation (EXP-14); profit and tonnage values are hidden — Table 1. Performance metrics of optimised plans for selected experiments. Confidential information has been removed. Open Table 1 full size

In our view, the most noteworthy finding from this study is shown in experiment EXP-14, where two new products replace all five existing export products (Product Consolidation). This result displays a considerable profit increase of 14%, yet also presents a remarkably simplified marketing strategy.

Results analysis suggests that the observed profit improvement most likely derives from a better alignment between product definitions and the Ash-CV-Yield relationship of the mined coal.

Improvements to market strategy discovered in these experiments typically corresponded with modest increases in the total volume of unsold coal. If model constraints forced the sale of all ROM coal, this affected the findings of this study, particularly for experiments that alter the existence of low energy product definitions. This conclusion is not surprising in the context of this study, as current market conditions were generally known to be unfavourable for certain plies.

Additional experiments (not shown) indicated that quantitative results and conclusions are sensitive to cost modelling assumptions, particularly as costs approach parity with revenue for select product definitions. This is an expected outcome, but it highlights the importance of accurate financial modelling, particularly under tight market conditions.

The key insights summarised in this study were unexpected and could have considerable impact for the client. Before communicating these results to a wider stakeholder audience, the client used an independent third party consultancy to validate the findings. The client also used its BlendOpt software licence to extend the study to include:

a more comprehensive cost model, to validate the robustness of the reported results to different cost assumptions
evaluation of result sensitivity to alternative price forecast scenarios
replication of the experiments on additional mine schedules, using relevant marketing planning time horizons.

Technical detail

Study methodology

New product modelling

As shown in Figure 2, pricing discounts for Newcastle indexed thermal products (e.g. 5950, 6250, 6200, 6350) exhibit a strict linear relationship with product CV acceptance ranges. Any interpolated thermal export product is therefore expected to follow the same linear function shown in Figure 2.

Figure 2: scatter chart of price discount against minimum product energy (CV) for NEWC-indexed products, with current and new product definitions falling on the same linear trend line — Figure 2. Price discount as a function of minimum product energy for NEWC-indexed products.

Other quality attribute ranges (moisture, VM, S) were derived in this study by linear interpolation of the two current products most similar in energy. This “locally linear” assumption is a close approximation to the quality targets a marketing team would negotiate (analysis not shown). Moreover, small deviations from these assumptions are expected to have a trivial impact on the final results, because these quality attributes are rarely active constraints in the customer’s optimal marketing plans (results not shown).

Definition of new products

Experiments restricted to the existing product definitions produced optimised results with consistently high volumes of product 5950 and consistently low volumes of the minimum energy (5700) and maximum energy (6350) products. An indicative example is shown in Figure 3.

Figure 3: bar chart of eight-month product volumes in a typical optimised plan, dominated by product 5950 at 10.2 Mt, followed by domestic coal at 1.05 Mt, 6250 at 714 kt, 6350 at 694 kt, 5700 at 310 kt and 6200 at 97.5 kt — Figure 3. Eight-month product volumes within a typical optimised plan.

Given the high volumes of product 5950 and low volumes of 5700, we anticipated that product strategy improvements were unlikely to involve new products with a CV lower than 5950’s specification. Most tests in this study therefore involved new product definitions with CV ranges between the 5950 and 6350 products, namely the new 5800 and 6300 products shown in Table 2.

Table 2: export product specifications, with NEWC price discounts and minimum and maximum moisture, ash, volatile matter, CV and sulphur, for existing products 6350, 6250, 6200, 5950 and 5700 and the new products 5800, 6100 and 6300, shown in red — Table 2. Export product specifications (new products shown in red). All data values, labels and discounts have been significantly obfuscated to protect the client. Open Table 2 full size

To reduce the number of new products to consider, we used an equidistance interpolation to define any new products (see Figure 2), which greatly reduces the number of new products that can be defined. In particular, new product 6100 is an exact equidistant specification between products 5950 and 6250, while new product 6300 is equidistant to products 6250 and 6350. Some tests also involved new product 5800, which was created by extrapolation of products 5950 and 6250.

Test coverage

Given the time-boxed nature of this study, a systematic Design-of-Experiments approach to test coverage was not used. Instead, the test coverage shown in Table 3 was generated with an exploratory procedure, in which each new test condition was motivated by the findings of previous tests. Only selected results from the tests below are presented in this report.

Table 3: matrix of the products included in the baseline and in each experiment from EXP-1 to EXP-14, across existing products and the new 5800, 6100 and 6300 products — Table 3. Product combinations for each experiment. Open Table 3 full size

Hypotheses

Hypothesis 1: The current set of thermal export product definitions is optimally aligned with the mine site’s distribution of coal energy, ash, washability and associated mining schedule. Evidence that potentially contradicts this hypothesis is found in the distribution of product volumes (Figure 3), as well as additional results analysis (not shown).

Hypothesis 2: A smaller product set cannot replace the existing product set while achieving a financially competitive result for the client. Evidence that potentially contradicts this hypothesis is found in the distribution of product volumes (Figure 3) as well as the linear relationship between price and CV (Figure 2).

FAQ

Coal technical marketing optimisation FAQ

How BlendOpt tests coal product definitions against what the mine produces and what the market pays.

What is technical marketing optimisation for coal?

It is the use of optimisation to test whether a mine’s coal product definitions are well aligned with the coal it produces and the markets it sells into. BlendOpt values each candidate product set against the mining schedule, washability, CHPP and logistics constraints, sales contracts and price forecasts. This lets technical marketing build a business case for product changes that operations can actually deliver.

How can changing a product specification increase profit?

In this study, replacing product 5950 with a new product 6100 required a 150 kcal/kg (2.7%) increase in minimum CV. BlendOpt found an optimised product strategy with an 8.4% increase in total profit over eight months. The analysis suggests the gain came from better alignment between product definitions and the Ash-CV-Yield relationship of the mined coal.

Can fewer coal products be as profitable as more?

Not always, but in this study two new product definitions replaced all five existing export products with an eight-month profit improvement of 14.3%. This is notable because fewer unique products are usually expected to reduce the profitability that product diversity offers. A smaller product set also brings practical benefits through simpler operations and business processes.

How are prices set for new coal products that are not yet traded?

For the Newcastle indexed thermal products in this study, pricing discounts had a strict linear relationship with product CV acceptance ranges, so new products were priced on the same linear function. Other quality ranges such as moisture, VM and sulphur were interpolated from the two current products most similar in energy. New products were defined by equidistant interpolation between existing products, which kept the number of candidates manageable.

How reliable are the results of a technical marketing study?

The results depend on the accuracy of the financial model, and this study found they were sensitive to cost assumptions where costs approach revenue for some products. Before sharing the findings more widely, the client used an independent third party consultancy to validate them. The client also used its BlendOpt licence to extend the study with a more comprehensive cost model, alternative price forecast scenarios and additional mine schedules.

Technical Marketing Optimisation Software for Coal