Northeast Energy Markets: Windfall or Washout?

Fortnightly Magazine - January 1 2000

Weighing the outlook for new plant investment in gas-fired power and related infrastructure.

The jury is still out on the type and size of additional energy infrastructure desirable in the Northeast United States, but enough data is in to make a few guarded observations.

The situation is fluid. Last June, the Federal Energy Regulatory Commission held a one-day public inquiry into anticipated demand for natural gas in the Northeast United States, eliciting a mixed bag of predictions.[Fn.1] In fact, just a few months before, FERC Commissioner William Massey had expressed confusion with a move by the Maritimes & Northeast Pipeline to downscale its capacity by 20 percent, calling it "counterintuitive."[Fn.2] But then again in late 1998, the FERC had asked the New England Power Pool to file a new plan on transmission access and grid constraints, on finding it unlikely that all new generation projects proposed for the region would actually be built.[Fn.3]

Nevertheless, despite the lack of institutions for a well-integrated power market in the Northeastern United States, the power industry is forging ahead with ambitious plans for new natural gas power generation there.[Fn.4] In this same region, the industry is pushing to build natural gas infrastructure that would serve all sectors on par with corresponding markets in most other parts of the lower-48 states, despite enormous regulatory and political hurdles.[Fn.5]

These efforts offer many riches at first glance, but are still fraught with peril. On the downside, natural gas retailers and wholesalers are losing residential market share relative to their counterparts on the oil side of the energy business. Poor economic and population growth in New York and nearby states has not dissuaded companies from planning huge infrastructure investments there.

On the upside, however, there is great potential for growth for residential gas sales in the Northeast, plus much opportunity in power markets. Many older, less-efficient generating plants should be up for retirement soon, opening a window for new gas-fired generation (and opportunities, perhaps, for distributed generation). Constraints in the region's transmission grid also adds value to gas resources and pipeline capacity.

The data reviewed here should provide not only a background on the situation in Northeast energy markets, but on why developers are even considering all the investment now planned in the region's energy infrastructure. Armed with information, we can snoop around and seek out where the highest returns might be found and whether they are likely to continue.

Residential Gas Demand:

Depressed by Low Fuel Oil Prices

Natural gas distribution companies have not shown much competitive mettle in the last several years in their sales to residential customers, the major natural gas-consuming sector in the Northeast. Natural gas as a percentage of total natural gas and distillate fuel use - the major space heating fuels in the Northeast - declined by 6 percentage points in New York between 1996 and 1998. If the natural gas share had not fallen, gas consumption in New York, which accounts for 66 percent of total residential consumption in this region, would have been 28 billion cubic feet (Bcf), or 77 million cubic feet (MMcf) per day, greater. Since both oil and gas demands are influenced by weather, this decline in share is not explained much, if at all, by the warmer weather in 1997 and 1998.

Natural gas accounts for less than 50 percent of the combined distillate oil/natural gas residential market in practically every Northeastern state except New York.[Fn.6] Although the trend of decline should be unsettling to the gas industry, the low percentages, especially when compared to the overall average natural gas share of about 88 percent for the rest of the United States, provide some idea of the potential for growth in this region. If the share grew to 88 percent in the region, then regional demand would increase by about 300 Bcf, or about 800 MMcf per day.

Part of the gain in market share by oil in recent years is due to price. Computing the fuel oil price relative to the natural gas price (see last column of Table 2) is a way of assigning a competitive score to gas distributors in the Northeast. Examination of these scores with 1998 data reveals that gas distributors in most states received a score equivalent to a letter grade of D. Only Maine and Vermont had a score of B or higher.

According to 1998 statistics from the U.S. Energy Information Administration (EIA), the difference between the prices distribution companies and heating oil dealers offered residential customers is a huge $3.00 per million British thermal units (Btu) in such states as Connecticut, Rhode Island and Massachusetts. The majority of consumers pay 40 percent more than the national average natural gas price of $6.49.

Power Demand: Depressed by High Prices

Residential customers in the Northeast not only pay much more for natural gas than the average customer in the remainder of the United States, but they also pay much more for power - 73 percent more in New York. In particular, customers in New York pay almost 6 percent of the revenues received by electric utilities from residential customers in the United States, yet they receive only 3.5 percent of the generated power. Not too surprisingly, electricity sales per residential customer also generally are much less in the Northeast than they are in other areas of the country. Accordingly, generation capacity per capita and use per customer are both low because prices are high. Thus, low figures for generation per capita in the Northeast do not necessarily indicate a lack of generation, as it might appear.[Fn.7]

External Factors: Slow Growth Continues

for Population and Regional Economy

About one-third of the counties in New York experienced a decline in population between 1990 and 1998. Counties in the other Northeastern states also experienced declines.[Fn.8] More recently, the population of the United States increased by almost 1 percent between 1997 and 1998. The only state in the Northeast that exceeded this national average was New Hampshire. Massachusetts and Vermont were next in growth, with 0.5 percent and 0.4 percent growth, respectively. The remaining states had a gain of less than a quarter of the national average.

The average real economic growth rate for the United States, as measured by real gross state product, was 20.5 percent between 1990 and 1998. Only two states in the Northeast - Vermont and New Hampshire - exceeded this percentage. New York's growth rate was a paltry 12 percent, while growth rates for Rhode Island, Connecticut, Maine, Massachusetts and New Hampshire were near 18 percent. Except for Maine and perhaps Connecticut, economic growth rates in the Northeast United States for the next several years are expected to be much less than the national average.

Since economic and demographic characteristics do not support strong market growth, especially for the residential sector, we next look at the potential for growth in terms of industry targets based on overall industry achievements.

Untapped Potential:

Room for Growth in

Residential Gas Demand

In most parts of the United States, especially where low winter temperatures and high space heating loads prevail, natural gas serves the space heating and water heating needs of a large percentage of the households. And in the Northeast, especially, there is much room for growth before gas market penetration and saturation reach levels observed elsewhere.

Because of its convenient and reliable use, natural gas often has quickly penetrated residential energy markets, especially space heating markets when the needed infrastructure was put in place, even if natural gas was slightly more expensive than the fuel in use for this service. In particular, the number of households, as reported by the U.S. Census Bureau, is highly correlated with the number of natural gas customers in those states that have extensive natural gas systems in place. In such states, the percentage of households that are gas customers, or the penetration, often is near 71 percent, and the percentage of gas customers that are space heating customers, or the saturation, is near 99 percent.[Fn.9]

Table 3 presents results for two targets. These targets represent the additional natural gas volumes in the Northeast that would result from penetration (Target 1) and saturation levels (Target 2) reaching national levels. As shown in the final entry in the table, the additional load from growing the market from the current level of penetration to 71 percent, and from the current level of saturation to 99 percent is 340 trillion Btu, which is about 931 MMcf per day.

To understand this result, first use 83.2 MMBtu per year as an estimate of space heating load per customer.[Fn.10] Then using New York data, simply multiply 4,117,000 customers by 83.2 MMBtu to obtain 342,534 billion Btu of possible space heating load per year. Next multiply this number by 13 percent (99 percent - 86 percent) to get 44,529 billion Btu, an estimate of the additional load from saturation reaching national levels in New York.

For estimating the total load of new customers, first add 30 MMBtu per year, a conservative estimate of non-space heating loads per customer per year, to 83.2 MMBtu. Then simply multiply 6,749 by 113.2 MMBtu and then by 10 percent to get 76,399 billion Btu, an estimate of the additional load from penetration reaching national levels in New York. When new customers are added, it is assumed that they will use natural gas for both space heating (83.2 MMBtu) and non-space heating services (30 MMBtu), and hence will contribute 113.2 MMBtu of additional load.

Old Power Plants: Replacement

a Net Positive for Gas

A recent EIA study pinpointed that non-fuel costs of generators built before 1960 were, on average, about four times as great as for generators of later vintages.[Fn.11] Naturally, this suggests that newer generators might replace these older generators in competitive markets. The total number of pre-1960 oil and gas generators in the Northeastern states represents significant amounts of generation.

As of 1997, New York had 30 power plants built before 1960 capable of switching between oil and gas, totaling 1,373 megawatts of nameplate capacity. Massachusetts and Connecticut had a total of 29 oil and gas units representing 2,186 MW of nameplate capacity. To the extent these older units are currently burning gas, their replacement with newer, more efficient units with higher heat rates implies a probable decrease in natural gas demand.

By contrast, as of the same date, New York had 1,946 MW of nameplace capacity in pre-1960 oil- and coal-fired power plants. Moreover, Vermont, Maine and New Hampshire had 37 old oil-fired plants, totaling 656 MW in nameplate capacity. Replacement of these units carries different implications. In fact, whether these units even represent likely candidates for direct replacement by new gas-fired plants depends on whether they have access to gas infrastructure. If they are not well-connected to gas supply lines, they may prove likely targets for new investment in gas infrastructure.

Thus, the likely retirement of older plants in this second group of facilities may represent potential growth - not only for the gas commodity, but also for capital investment. Certainly it would appear to offer at least regular returns for new investment in gas-fired power in the Northeast. This conclusion holds up under a price comparison between gas and electricity. In fact, an examination of power and wholesale natural gas prices for the last several years (see Figure 1) reveals that natural gas prices, in equivalent units using a heat rate of 7.8 MMBtu per megawatt-hour (MWh), almost always are less than power prices.

Price Volatility: Also Enticing

for Gas-Fired Power

The power price volatilities in the Northeast are enormous at times and under ordinary conditions exceed natural gas price volatilities, which already are huge when compared to those of other traded commodities. In addition, large changes in wholesale power prices in the summer elicit little or no change in gas prices.

High price volatilities and low correlation between changes in power and gas prices may explain much of the interest of some companies in obtaining rights to operationally flexible gas generation in the area. In fact such situations are dreams come true for companies with effective trading and hedging arms able to keep close track of the market. These companies are in the gravy as long as the conditions supporting the high volatilities and low correlations continue.

Pipeline Prices:

Capacity Release Market

Carries Promise of Discounts

Probably the most important pipeline routes for moving gas into the Northeast are Tennessee Zones 5 and 6. Interestingly, capacity release markets in both zones have been quite robust.

The size of discounts per transaction available was sizable during the more than 5,000 capacity release transactions for the period 1996 through 1998 on Tennessee Zones 5 and 6. The median value indicates that for 50 percent of the transactions, discounts of 71 percent and 87 percent off normal rates were received on Tennessee Zones 5 and 6, respectively. The statistics also indicate that 50 percent of the capacity release transactions on Tennessee were for 558 MMBtu per day or less. That figure represents just about the amount of gas needed to produce 5 MW of power per hour from a generator with a heat rate of 7.8 MMBtu per megawatt-hour, operating for 16 hours.

The large discounts available on Tennessee should provide opportunities for shippers to move gas to power generators during the non-heating season since the largest number of releases to both zones occurred during the spring and summer months. Although the low price may suggest excess overall pipe capacity, that is not necessarily the case. Why is that? Sellers are willing to sell the capacity at a low price because of the high insurance premium associated with their retaining these rights for certain times of the year or for the future, when demand grows or is likely to surge. At such times, these rights have great value.

The Power Grid: Constraints Boost Value

of Gas Pipeline Expansions

A close look at Figure 1 indicates that large changes in power prices in one region are not always coupled with large changes in price in another region. The existence, at times, of weak relationships between changes in power prices in the two areas suggests possible constraints in moving power between locations. However, the average wholesale power prices displayed in the figure are not comprehensive. Thus, they are not sufficient guides for indicating the implications of constraints in power transmission and distribution on power costs, especially under changing market conditions. Models that attempt to replicate the operation of the power systems are a better choice.

When performance of the New England region (NEPOOL, or New England Power Pool) and New York Power Pool (NYPP) is evaluated through a detailed model representing the operation of the system, some striking results are revealed.[Fn.12] When the system is near full capacity in New England, under the regulated regime, the marginal cost is double that in New York. Nearby areas have much different marginal costs. This situation suggests that companies with power generation in the poorly connected areas will obtain significant profits, especially when demand surges.

The figure also reveals that under deregulation, when demand in areas south and southwest of the Northeast is reduced by as much as 20 percent, and thus capability to export power to markets in the Northeast is improved, not that much happens in some areas.

City Gate Prices: High Retail

Markups Should Fall

In today's gas industry, there is a price dance in which wholesale spot prices are played against city gate prices, and city gate prices are played against end-use prices for energy services. The sport involves sizing up the differential cost of different delivery services. As exposed in the review of capacity release markets, discounts for transportation service may be huge when the expected use of the system is much less than full capacity.

The first item to note in Table 5 is the spread between retail commercial gas prices and city gate prices. These spreads are enormous because the reported commercial price represents the price paid by on-system customers. This price does not represent the average cost to all commercial customers both on- and off-system, since an increasing number of large commercial customers now receive gas via non-utility providers, generally at a lower price.[Fn.13] Next note the difference between city gate prices and wholesale prices. Average wholesale prices are generally less than city gate prices. That tends to put downward pressure on city gate prices and, as a consequence, on retail prices as well.

Many off-system providers of natural gas attract commercial customers by offering a price lower than the retail price of the incumbent utility. That pressures distribution companies to lower their price to the remaining customers. In fact, the cost of gas to on-system customers in Massachusetts has fallen each year, from $6.88 per thousand cubic feet in 1996 to $6.08 per thousand cubic feet in 1998. Not surprisingly, the number of off-system customers has risen each year, from 58,135 in 1996 to 157,319 in 1998. They now comprise almost 50 percent of the commercial customers in Massachusetts.

On-Site Power: Benefiting

from High Spark Spreads

To calculate the expected savings in New York from using natural gas to generate on-site power, first choose a particular generation unit and its power-using characteristics. The unit chosen as an example is a distributed generation (DG) unit capable of producing 100 kilowatts of power per hour. The purchaser of this unit would be a fairly large commercial establishment, such as a large restaurant or retail store. Then assume this commercial establishment operates, on average, 12 hours per day, six days per week. That amounts to 7,200 kilowatt-hours (100 kW x 12 hours per day x 6 days) or 7.2 MWh each week. If we multiply this amount by $79.18 (New York's spread of $89.18 less a non-fuel operating cost of $10.00 per megawatt-hour, which includes all variable costs from hookup to regular maintenance), we obtain a value of $570. When this amount is multiplied by 52 weeks in a year, annual savings of $29,645 are obtained.

The cost of a 100-kW unit is between $90,000 and $100,000. Thus, the simple payback period is a little more than three years. Naturally, the more hours and the greater the number of days of operation, the better the economics. On the other hand, if this unit is used at less than full capacity, the poorer the economics.

If the best price the commercial customer was able to negotiate was the on-system retail industrial customers' price - $4.09 per thousand cubic feet, or $3.97 per million Btu, in 1998 - then the spread would be $65.31 and total savings would be $24,452 instead of $29,645. The payback period would be about four years instead of three. It is worth noting that the off-system industrial customer probably pays less than $4.09 per thousand cubic feet.

The economics work out similarly for units with capacities from 30 kW to 200 kW because the cost per kilowatt for units of this size is $900 to $1,000 per kilowatt. Finally, if the system were only operated, on average, at 80 percent of full capacity, then the savings would be 80 percent of $24,453, or $19,532, over a payback period of five years.[Fn.14]

Overall, the loads of large commercial customers likely will be relatively constant, or if not constant, may even increase in the summer to satisfy a space-cooling load previously satisfied by their local power providers. Since the gas system generally is underused during the summer, that will improve the overall economics of the gas system. Fixed investment will be spread out over an increasing amount of gas. If enough of these customers are added, then the average cost of gas per customer in the summer will decline for all customers. That potential provides an additional incentive for distribution companies to lower the cost to these customers and an additional reason for these commercial customers to negotiate a reduced transmission/distribution cost from the utility.

Infrastructure Outlook: A Mix of Factors

The major growth area for natural gas in the Northeast United States during the last five years has been the commercial sector, and that is expected to continue through 2000. Even without including the impact of the new Maritimes & Northeast pipeline, natural gas demand in the commercial sector is estimated to increase by at least 200 MMcf per day between year-end 1998 and year-end 2000, using average growth rates experienced in the 1990s.

Increased growth in this sector will continue to be influenced by price reductions obtained by customers from non-utility providers. Combined demand in the utility and industrial sectors also is expected to increase by about 190 MMcf per day. However, forecasts for these sectors may be understated because of reporting problems with the historical data used to obtain the forecasts.[Fn.15] The utility sector forecast also does not include any substitution of new gas generators for pre-1960 generators, because the most recent data indicates this replacement has not yet begun.

The potential demand from the residential sector, which consumes more natural gas than any other sector in the Northeast, appears to be great. Recent EIA forecasts indicate that distillate oil prices will continue to be much less than natural gas prices for residential customers in the Northeast and New England regions in the next several years. Yet, the difference in prices is expected to become smaller.[Fn.16] Although demand will increase because some households will be willing to pay the reduced premium price associated with the more convenient source of energy, demand will not change much unless price declines significantly. But if price does decline significantly, the results could be striking and not without precedent.

Between 1964 and 1969, saturation in New England increased from 34 percent to 46 percent, or an annual percentage change of 6.2 percent.[Fn.17] If such a change were to occur today, residential demand would increase by 167 MMcf per day within two years. However, such progress depends on elimination of restructuring fees and other conditions that make it difficult for outside companies to compete with the incumbent utility and provide customers with a lower price for natural gas.[Fn.18]

Interestingly, the commercial sectors and high-income households in the Northeast appear quite promising as a market for distributed natural gas power generation.[Fn.19] In fact, increased natural gas pipe capacity in conjunction with natural gas distributed power generation may turn out to be a possible means of circumventing the enormous institutional, economic and political problems associated with creating an integrated, economic and reliable power market in the Northeast. If commercial, large-residential and small-industrial customers use increasing amounts of natural gas for DG, then there will be less need to trade wholesale power across power markets. Gas will be traded instead.

As DG increasingly is used, there also may be fewer instances of wholesale power price surges and limited transmission capacity. Since natural gas markets, especially in the East, are becoming increasingly well-connected, there may well be an overall improvement in the efficiency of energy markets and less volatility in wholesale power markets, but some increase in the price of wholesale natural gas. However, there also will be less value associated with new investments in conventional, operationally flexible natural gas power generation, and less will be installed.[Fn.20]

John H. Herbert is adjunct professor of statistics at the Northern Virginia Graduate Center of the Virginia Polytechnic Institute & State University. He also provides expert testimony and consulting service on such topics as fundamental industry analysis, applied econometrics and forecasting, quantitative methods for asset valuation, price risk management and the interrelationship between commodity markets and price volatility. Before opening his private consulting practice, he served as a senior economist at the U.S. Energy Information Administration. He may be reached at jhh1@msn.com.

1 See "Gambling on Gas Demand: Timing is Everything," by Richard Stavros, Public Utilities Fortnightly, July 15, 1999, p. 54.

2 See "News Digest," Public Utilities Fortnightly, June 1, 1999, p. 11.

3 See "News Digest," Public Utilities Fortnightly, Jan. 15, 1999, p. 14.

4 Energy Information Administration in its Form 860B indicates that there are 6,072 MW of capacity planned for non-utility generators for 1999-2003. Because of confidentiality concerns, EIA does not report planned utility capacity additions for Northeast Power Coordinating Council (NPCC), which is very similar to the definition of Northeast used here. Our estimate based on reported information for other locations and aggregate information is 300 MW, which provides an estimate of about 6,400 MW of planned capacity between 1999 and 2003. Steven H. Watts, McGuire Woods Battle & Boothe LLP, www.mwbb.com/services/energy-mp.html, in early November indicated 2,178 MW is under construction, 1,352 MW is under development and 19,656 MW is proposed. Energyinsight, RDI, in an Aug. 16 article indicated that developers intended to propose 9,100 MW in New York with nearly 5,000 MW designated for sites around New York City. Energyinsight on Aug. 23 indicated that 25,385 MW of planned merchant plant capacity is likely to be online in NPCC by 2003.

5 Approved projects such as Millenium from Ontario into New York, CNG into Empire and Tenneco Link to an LNG facility, could add more than 800 MMcf per day of capacity by 2002. Approval of the 682-MMcf per day Transco Market Link stalled in a lawsuit brought by Gov. Whitman in New Jersey. Planned, proposed and announced projects such as Duke Energy's Crossbay, Iroquois from Quebec into New York (East Chester), Tennessee's Eastern Express into Connecticut, Iroquois expansion into Vermont and Algonquin's Hub Link could add an additional 1.4 Bcf per day. Pipeline Tracking Database, Natural Gas Division, EIA, USDOE.

6 For the commercial and industrial (C&I) sectors, the natural gas share of the combined oil (which includes distillate and residual fuel oil) and gas consumption is much higher than for the residential sector. The C&I gas shares exceed 75 percent and 80 percent, respectively, while the residential share is near 50 percent. The gas share far exceeds the oil share for all states except Maine, New Hampshire and Vermont for the C&I sectors.

7 The average sales per residential customer in New York and Massachusetts is 8,123 kWh and 6,626 kWh, respectively, while for the average U.S. customer it is 10,371 kWh, "Energy Sales and Revenue," 1998, EIA. Electricity consumption in the Northeast has only increased at an annual rate of 1 percent during the 1990s. EIA projections for the summer of 2000 report a capacity margin (capacity relative to expected sales) of 17.2 percent for NPCC, the second-highest among the 10 NERC regions. These statistics represent the sum of each utility's capacity relative to expected sales and do not appear to account for such changes in business practice as the increasing amount of sales of power to customers supported by power transactions between utilities, which can, in turn, be supported by the more intensive use of existing capacity. "Electric Power Annual 1997, Volume II," October 1998, p. 69, EIA.

8 Between 1990 and 1998, the population loss from New York (excluding immigration from outside the United States) was 195,000. This number was more than twice as large as the next largest out-migration state, California, which has 80 percent more people. Interestingly, New York and California generally had the highest end-use energy prices in the Eastern and Western United States, respectively. All other things being equal, high energy prices necessarily will drive companies, and hence workers, out of the state. For examples of the relatively high natural gas and power prices in these states, see John H. Herbert, "The ABCs of Trading BTUs," Energy Intelligence Group, New York, N.Y., May 1998.

9 According to EIA's "Residential Energy Consumption Survey," penetration and saturation for New England in 1997 was 35.8 percent and 78.9 percent, respectively. According to the American Gas Association, estimates derived here for 1998, we get an estimate of 38 percent for penetration and 78.6 percent for saturation.

10 This estimate was obtained from "Residential Energy Consumption Survey," 1997, EIA. This report provided a separate estimate for customers in New England. Separate estimates for New York are not available. The lowest level of aggregation that includes New York is the Mid-Atlantic, which includes Pennsylvania, New Jersey and New York. The New England estimate is used for New York here.

11 "Issues in Midterm Analysis and Forecasting," 1999, EIA.

12 Overbye, Thomas J., Douglas R. Hale, Thomas Leckey and James D. Weber, "Assessment of Transmission Constraint Costs: Northeast Case Studies," to be presented at the IEEE Winter Meeting in February.

13 Non-utility providers often negotiate and pay a charge to use the utility's distribution system to serve the customer, but they do not provide information on the negotiated price to the EIA. However, the EIA obtains information on the volumes and the number of customers because the utility still delivers the gas to customers.

14 For a detailed economic evaluation of distributed resources see "Information to Support DR Business Strategies: Quantitative Analysis of DR," EPRI.

15 It has been increasingly difficult for EIA to obtain comprehensive and accurate data for off-system sales especially for large industrial customers because of the wide variety of arrangements used to arrange for receipts of gas. In addition, when utilities sell off generation, the sold generation is classified as non-utility or industrial-sector power generation. Thus, the use of natural gas by such generators should, in principle, be included in with electric utility consumption until they are sold (most of the generation sales in the Northeast occurred in September 1998 and in June 1999.) After they are sold, the use of natural gas by such generators should be included in the industrial sector. Obviously, the data accounting doesn't capture this well. In addition, the change in accounting makes the use of most recent data as a basis for forecasting most difficult. In the Northeast, 7,328 MW and 8,559 MW of utility plants in 1998 and 1999 (as of June 1999) was reclassified as non-utility plants. This is particularly important because these plants are major growth sectors for natural gas and power generation. These generators also have a greater incentive to operate longer hours, seek out profit opportunities and hence use more natural gas than otherwise similar plants. Thus, this change in terms of obtaining accurate forecasts also is of great importance.

16 "Annual Energy Outlook," 1999, EIA. The difference in the oil and gas prices declines over time, but the decline in the most recent forecast volume (for 2000) is less than in the previous forecast volume because end-use sector restructuring is progressing more slowly than expected.

17 For a discussion of matters concerning growth in residential markets, see John H. Herbert, "Clean Cheap Heat - The Development of Residential Markets for Natural Gas in the United States," Praeger, New York, N.Y., 1992.

18 Significant changes have already occurred in the C&I sectors, and that bodes well for residential sector. Sales of natural gas to C&I sectors are much higher than sales of oil, and an ever-increasi

Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.