# The economics of a chemical plant: costs, prices and profit: Markets: sizing demand and forecasting price

Source page: https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/p2/

[Understanding Chemicals](https://www.materialinsights.org/learn/chemicals/) / [Module 3: The economics of a chemical plant: costs, prices and profit](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/) Part 2 of 5

## Markets: sizing demand and forecasting price

Sensitivity studies keep finding selling price and sales volume to be the inputs that move project returns most, yet both must be forecast years before the first sale. This part covers the main ways to size a market, how a price is built from a floor cost and a competitive margin, and why forecasts so often fail together.

### 2.1 Market size is estimated from end uses, trends, economic links and other countries

Brennan offers four ways to estimate volume and advises using several, because each is imprecise.

End-use analysis starts from the customer industries. For low-density polyethylene he assumes 34% goes to film and sheet (26% packaging, 8% construction, household and farm uses), 26% to injection moulding and 40% to other uses such as cable insulation and coatings. Each segment is sized, the share LDPE holds is estimated, and future penetration is forecast against competing materials. Field surveys are costly, but errors in different segments tend to cancel, so the method is reliable.

Statistical projection fits a trend to past consumption and extends it. Products tend to follow an S-shaped life cycle: a slow initiation phase, rapid growth as bigger plants cut costs and lower costs widen the market, then maturity. Logistic or Gompertz curves describe the S; industry shorthand is an exponential growth rate. The hard part is knowing where on the curve a product sits.

Correlation ties demand to economic aggregates such as GDP, capital investment or industrial production, or to one sector: PVC demand follows building activity. Per capita comparison looks at consumption in more industrialised countries and asks how fast a developing country may follow.

Growth also stops. Brennan lists the brakes: high cost or price, technical limits, limits to equipment scale, and health or environmental concerns, which in the extreme ended the market for chlorofluorocarbons in the 1990s.
 Case

#### Sodium cyanide in Australia

Sodium cyanide, used to leach gold from ore, shows how hard demand can be to foresee. Australian imports were about 2,000 tonnes in 1975. Gold prices roughly doubled in real terms between 1975 and 1988, Australia's share of world gold output rose from 1.4% to about 10%, and imports reached 52,000 tonnes in 1988. Local plants started around 1990. By 2018 Orica at Gladstone had 95,000 tonnes a year of capacity and Australian Gold Reagents at Kwinana 78,000. Brennan judges that 1990 demand would have been very hard to forecast in 1975.

- Brennan (2020), Ch 2 Market evaluation and forecasting · pdf pp. 24, 30-34, 53-54

### 2.2 Price is a floor set by the best cost, plus a margin that competition erodes

Brennan, following Rudd and Watson, splits a selling price into two parts. The floor price is the lowest price that covers production cost with the best available technology and scale, plus a minimum return on capital. The margin is whatever extra the market allows. A new product can carry a large margin to repay research and risk; as rivals arrive, the margin shrinks. Both parts fall in real terms: the floor through better technology, larger scale and learning, typically by 1-4% a year, and the margin through competition, typically by 10-20% a year.

In a glut, suppliers cut towards the floor and may briefly sell below it; in a shortage the margin widens.

Where imports are allowed, the ceiling is import parity: the lowest price at which imported product can be landed. Brennan builds one for a polymer shipped from Europe to Australia. The FOB price (free on board, the price loaded on the ship at origin) is 1,200 euros a tonne; duty at 10% of FOB adds 120, freight 160 and landing charges 20, for 1,500 euros, or A$2,400 at 1.6 Australian dollars to the euro. Exchange rates therefore move the ceiling, which makes it hard to forecast.

Some products escape import competition. Liquid chlorine is too toxic to ship far, and for some products transport costs are too high relative to value to ship them over long distances, which shields local plants. Their derivatives may still face import competition.

Data Import parity price for a polymer shipped from Europe to Australia

| Element | Euros per tonne |
|---|---|
| FOB price in Europe | 1,200 |
| Duty at 10% of FOB (hypothetical tariff) | 120 |
| Freight (say) | 160 |
| Landing charges (say) | 20 |
| Landed import parity price | 1,500 |
| In Australian dollars at A$1.6 per euro | A$2,400 |

A local producer cannot sustain prices above this ceiling while imports are free to enter.
 Source: Brennan (2020), Ch 2 Market evaluation and forecasting, pdf pp. 35-36  Worked example

#### How fast a margin disappears

Illustrative arithmetic within Brennan's ranges. A new product sells at a floor of $1,000 a tonne plus a margin of $500. If the margin decays at 15% a year it is multiplied by exp(-0.15t): after 4.6 years it has halved to $250, and after 10 years it is about $112. If the floor decays at 2% a year it is about $819 after 10 years. The real price has gone from $1,500 to about $930. A project priced at launch levels for life would be badly overvalued.

- Brennan (2020), Ch 2 Market evaluation and forecasting · pdf pp. 34-36, 39
- Brennan (2020), Ch 4 Operating cost estimation · pdf p. 120

### 2.3 Elasticity, lead times and the investment cycle make forecasts err together

Price elasticity of demand measures how strongly volume responds to price. Brennan defines it as the fractional change in volume divided by the fractional change in price. Demand that ignores price is inelastic; demand that falls away sharply when price rises is elastic. The main driver is the availability of substitutes. Glass beer bottles face aluminium cans, so their elasticity is high. A petrol car cannot switch fuel, so short-run elasticity for petrol is low, but over years people buy different cars. Long-run elasticity is larger than short-run.

Brennan notes 2-3 years from sanction to commissioning, then several more before demand fills a new plant. Over-optimistic forecasts led to worldwide petrochemical overcapacity in the early 1980s, falling prices and closures.

Wittcoff calls this the old paradigm: overbuilding, and building at the wrong time. Rising demand lifts prices and profits, firms invest, and because plants take years to build, the new capacity arrives as demand turns. Prices fall, plants close, and the overshoot in closures sets up the next shortage. He quotes 4-5 years for a full cycle. Small imbalances have large effects: in chemicals a shortage may be only a 2-3% shortfall. When three US crackers and plants elsewhere were damaged in 1994-95, profits jumped, then fell back once repairs were done.

The forecasting record shows the bias. World ethylene forecasts underestimated demand in the 1960s, then overestimated it in the late 1970s and early 1980s, because forecasters were guided by decades of rapid growth and underrated the effect of expensive oil.

Data Ethylene: growth rates and the forecasting record

| Period | What happened |
|---|---|
| 1940-55 | Consumption grew about 16% a year (130,000 tons a year by 1946) |
| 1955-73 | About 11% a year; forecasts of the 1960s underestimated growth |
| 1973 to early 1980s | Growth sharply slowed as crude rose from about $2 to $12 (1974) and $28 per barrel (1980); forecasts overestimated demand |
| 1985-95 | Growth restored to about 5% a year |
| 2016 | US ethylene about US$660 a tonne, global about US$1,000 |
| 2018 | World production 154 million tonnes, double the level 25 years earlier |

 Source: Brennan (2020), Ch 2 Market evaluation and forecasting, pdf pp. 33, 55-56  Case

#### Restructuring after the ethylene glut

Wittcoff describes the US response to ethylene overcapacity built up in the late 1970s. Between 1981 and 1986 capacity was cut by about 10%, and the industry shed 7-8% of its 1.15 million employees. Demand then caught up and 1987-88 was probably the most profitable period US chemicals had known. By 1989 firms were debottlenecking and building again, and by 1991 there was overcapacity once more. In Western Europe, new crackers pushed average operating rates down to about 83% of nameplate capacity in 1993-94.

- Brennan (2020), Ch 2 Market evaluation and forecasting · pdf pp. 33, 36-38, 55-56
- Wittcoff (2004), Ch 1 The chemical industry · pdf pp. 48-52

### See it in the table
[Al Aluminium](https://www.materialinsights.org/commodities/aluminum/)[Au Gold](https://www.materialinsights.org/commodities/gold/)[Cl2 Chlorine](https://www.materialinsights.org/chemicals/chlorine/)[C2H4 Ethylene](https://www.materialinsights.org/chemicals/ethylene/)[PVC PVC](https://www.materialinsights.org/chemicals/pvc/)

### Check your understanding

Using Brennan's import parity build-up, what landed price in euros does a polymer with an FOB price of 1,200 euros reach?
 1,200 FOB + 120 duty (10% of FOB) + 160 freight + 20 landing = 1,500 euros. The A$2,400 figure is the same price converted at 1.6 Australian dollars per euro.

According to Brennan, what most strongly raises the price elasticity of demand for a product?
 Where a substitute exists, as aluminium cans do for glass bottles, buyers switch when price rises, so demand is elastic. Where switching needs new equipment, as with car fuels, short-run elasticity is low.

In the Rudd and Watson model used by Brennan, which part of the price typically decays fastest in real terms?
 The floor falls slowly (1-4% a year) as technology and scale improve; the margin falls quickly (10-20% a year) as competitors enter.

[← Part 1: What makes the process industries different](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/p1/)[Mark read · Part 3: Capital cost: what the plant costs to build →](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/p3/)

## Other sections

- [Module overview](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/index.md)
- [What makes the process industries different](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/p1/index.md)
- [Capital cost: what the plant costs to build](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/p3/index.md)
- [Operating cost: what each tonne costs to make](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/p4/index.md)
- [Profitability: does the project earn its capital back?](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/p5/index.md)
- [Key terms, quiz and sources](https://www.materialinsights.org/learn/chemicals/plant-economics-costs-and-profit/review/index.md)