Distillate Inventory Behaviour in the USA

Distillate (heating oil and diesel) inventories fell steeply in the USA in the early eighties (see figure below) and have lost some of the pronounced seasonal effect associated in particular with heating fuel oil . Why have stock levels fallen and what has driven the changing seasonal pattern in the 1980s and 1990s? Below I explore some possible reasons.

Distillate inventories grew from 35 to 205 million barrels (mb) (annual averages) over the 1947-1980 period. As is shown in the figure a large build-up was seen between 1973 and 1980, followed by a 62 percent fall in inventory levels until 1989, after which stocks gradually increased. In the 1990s stock levels gradually rose to 125 mb in 1995, but 1996 saw a sharp fall to an uncomfortably low 106 mb, then in 1998 stocks surged to 137 mb.

The seasonal effect (stock draws in the February to April period and stock builds in the July to August period) has dampened considerably over time. A number of reasons have been suggested as explanations for these trends: a) rationalisation in the refinery industry; b) a move to relatively higher crude stocks; c) “just-in-time” inventory management; d) changing demand patterns; e) increasing environmental regulation, and; f) an increasing reliance on imports. Below, we consider these in more detail.

Following a steady build-up of refining capacity in the 1970s the removal of price and allocation controls in 1981 saw the industry enter a period of rationalisation. The number of US refineries declined from 324 in 1981 to 163 in 1997, with most of the closures coming in the early 1980s. The fall of inventory levels in the 1980s was mainly due to the drive to cut costs. Stocks held by refiners declined by about 31 percent from 49 mb in 1981 to 34 mb in 1989. The largest drop in inventories was seen at bulk terminals where stocks went from 114 mb in 1981 to 45 mb in 1989. Stocks at bulk terminals are a seen as a direct variable cost, in the order of 1 cent per gallon per month, while stocks held at refineries are needed for smooth operation of the system and hence reflect demand more directly, a point we return to below.

New investment went into upgrading existing facilities by adding secondary processing equipment, and this together with capacity creep helped to offset the impact of refinery closures. By 1997 the number of refineries was half that of 1981 but, total distillation capacity only fell by 17 per cent. Utilisation rates increased from 66 per cent in 1982 to over 90 per cent in the late 1990s and, while distillation capacity has declined, output has increased. Increased refinery complexity and utilisation rates have facilitated greater flexibility for refiners, and hence lower stock levels were needed to avoid stock-outs. Perhaps the greatest impact has been on the seasonal component of distillate inventories. Greater flexibility has meant that distillate output has fluctuated less, even though runs have increased, which is reflected in distillate inventory behaviour. Over longer time periods this effect is quite marked. In the 1990s inventory draws and builds have been smaller than in the 1980s, and stock builds are drawn out over a longer period.

Some have suggested that relatively higher crude stock levels have allowed refiners to meet incremental demand with higher crude runs rather than using product inventories. With regard to the ratio of crude to distillate stocks a large rise in the ratio was seen in the early 1980s. It increased from approximately 2.5 in 1983 to above 3 in 1990. However, it had dipped back to 2.5 by 1998. Also relevant is that in terms of days covered crude stock levels have been below those for distillates and motor gasoline throughout the 1980-98 period, with days covered falling for crude oil and the two petroleum products during the period.

With “Just-in-Time” (JIT) inventory management one sees a fundamental shift in the supplier/refiner relationships or supply acquisition practices that spread the benefits and the risks between the two parties. Lags associated with waterborne or pipeline shipments restrict refiners in adopting JIT relationships. JIT appears to be a loosely used term in the literature when the authors really mean cost-cutting practices, some of which would also be undertaken by a firm that wishes to pursue JIT inventory management. [The interested reader is referred to the Oil and Gas Journal (8 July 1996) for a critique of JIT adoption in the US refining industry.]

Demand for distillates peaked in 1976-77 at 4.1 mb per day (b/d) and gradually fell to 2.7 million b/d by 1982/83. This drop in demand was due to fuel switching and energy conservation. Following a fall in prices demand has recovered to about 3.4 million b/d. The demand for heating fuel and diesel fuel has not followed the same pattern. Peak demand of distillate is driven by winter heating oil, consumption of which has declined steadily, in part due to competition from cheaper natural gas. At the same time demand for diesel fuel has gradually risen. Falling demand has lowered the demand for stocks, indeed, the (mildly) U-shaped pattern of demand is in line with the pattern displayed by refinery-held inventories, and the changing proportions between heating fuel and diesel reduced the seasonal component of stocks.

Environmental regulation has also had an impact on inventory levels, and in particular it has reduced the fungibility of oil product stocks. The increase in the number of grades has required more segregation in storage. These factors have created an upward pressure on stocks which may in part explain the gradual rise in distillate inventories in the 1990s. Finally, US imports and exports influence stock levels, as both can be used to balance supply and demand without needing recourse to inventories.

A number of factors combined to reduce distillate inventory levels in the 1980s. The two main factors were refinery industry efforts to raise profitability combined with falling demand. In the 1990s, distillate stock levels have remained steady or increased slightly. In part, his may be due to the lagged effect of increased environmental regulations which have increased demand for storage. The declining seasonal effect is largely due to increased refinery complexity coupled with a changing (heating oil/diesel) demand composition and a greater use of international trade.

By: Robert Mabro




Latest Tweets from @OxfordEnergy

  • New OIES study: Turkey’s gas demand may be no more than 55–56 bcm/year by 2025 and 60–62 bcm/year by 2030… https://t.co/9T5YhOKhyR

    April 26th

  • Gas will continue to lose out against coal in Turkey's power generation, says new OIES study @AAEnergyNews https://t.co/hjKBERfzLw

    April 25th

  • New OIES study: Ankara looks to renewables and domestic coal to cut energy & mineral import bill https://t.co/4O8CXhtIVD

    April 25th

Sign up for our Newsletter

Register your email address here and we will send you notification of new publications, comment, articles etc. automatically.