Analysis

The primary objective of this analysis is to compile methodology and information to assess the potential impact of shuttle train rates on the local grain market. Although shuttle rates are the market phenomenon addressed in this analysis, methodology generally is applicable for developing market synopsis for other factors influencing local grain delivery patterns. The analysis assumes a profit maximizing goal for local producers in their grain marketing decisions. Although elevator loyalty and other qualitative factors may influence the producer delivery decision, it is assumed that the local grain delivery patterns will satisfy profit maximizing criterion in the long run.

The analysis has two fundamental components. The first component concentrates on decisions made by North Dakota farmers in marketing their grain. Spatial analysis is used to couple producer trucking costs with local market board prices to estimate producer delivery patterns. Grain production and draw area spans are used as quantitative measures in discussing the delivery patterns. The second component specifically emulates local processor position in the local grain market. An economic decision model illustrates the impact of changes in elevator rail rates and producer trucking costs on the relative competitiveness of local processors.

Producer Delivery Patterns

Estimation of producer delivery costs is a rudimentary, yet essential, task in understanding local grain marketing patterns. A spreadsheet-based mathematical model of the decision was created to allow for replication of this analysis by local decision makers. The model relies on rail rates, truck costs, and distance in estimating producer decisions for local grain deliveries.

The delivery scenarios for marketing grain in proximity to shuttle facilities are estimated by calculating and plotting a series of indifference points for the producer decision. An indifference point is defined by that distance at which the net price to a producer is equalized for two competing markets. The net price considered by the producer, in this model, is equal to the market price less the cost of delivering the commodity to that market. A series of these indifferent points defines the outlying edge of the draw area for the shuttle market. The point of indifference, is defined by this equation:

Where,

This equation is applied to a multitude of market combinations to estimate grain draw area boundary scenarios for 10 shuttle facilities considered in this analysis. The boundaries provide a base for identifying likely grain origin territory for the facilities. The share of grain the facility handles from within the boundary vary across commodities, facilities, and time, because production and market conditions are dynamic. Each facility also has an individual business plan and a goal for annual grain volume, which will affect its market decisions and influence local grain distribution patterns. The model and estimated draw area boundaries do, however, provide a base for discussing local grain delivery patterns and infrastructure.

Base Case - Wheat

An initial step in assessing the influence shuttle train rates will have on the local grain market is identifying the area affected by estimating draw area boundaries for facilities with access to these rates. Wheat accounted for almost half of the grain produced in North Dakota between 1994 and 1999, thus, it is the base case commodity for this analysis. A truck cost of $.0025 is applied in the base case. This cost reflects the cost of operating a semi-truck and trailer, as described in the previous section of this report. Minneapolis is the terminal market used for defining applicable rail rates. A shuttle rate is not published for the Minneapolis market, so the proxy shuttle rate was applied based on industry comments and shuttle train rates quoted for other markets. The shuttle rate is assumed to be $300 per car (9.1 cents per bushel) lower than the unit train rate. The boundaries estimated using these parameters form the base case scenario and provide an approximation of the affected market area. The sensitivity analysis developed later in this analysis illustrates a range of potential impacts and the affect of alternative market influences on local grain marketing patterns.

The base case shuttle draw areas are illustrated in Figure 14. The unique shape of draw areas for individual shuttle facility locations reflect rail rate relationships, road network, and truck costs represented in the producer marketing decision. Because commodity weights and rail rates are equal, the base case draw areas are applicable for HRS wheat and durum. These draw area boundaries were applied to county wheat production data to calculate the volume contained in each of the draw areas. The production densities for the commodities are illustrated in Appendix B.

HRS wheat production is most dense in the eastern half of the state, with Pembina and Cass offering the greatest densities among North Dakota counties. The area included in the 10 shuttle facility boundaries accounts for approximately 45 percent of the total North Dakota land area. The 10 draw areas encompass about 38 percent of North Dakota HRS wheat production and 39 percent of the state's durum production. The net bushel estimates the combined draw area totals and accounts for the approximate draw area overlaps among shuttle facilities, of 7 and 17 percent, for HRS wheat and durum production, respectively, and a 6 percent overlap of land in wheat draw areas.

The draw area for the Alton location offers the highest volume of total wheat among the shuttle facilities, with an estimated 23.4 million bushels. Sterling, Gladstone, and Berthold draw areas are estimated to contain more than 17 million bushels of wheat. Approximately 21.1 million bushels attributed to Sterling and 18.5 and 17.8 to Gladstone and Berthold, respectively. Draw areas for Kindred, Scranton, and Jamestown fall in the 10 to 15 million bushel range, with 14.7, 14, and 11 million bushels of wheat produced in their respective draw areas. The Bowbells, Northgate, and Bottineau facilities are encompassed by draw areas each estimated to include less than 10 million bushels of wheat.

Of these six facilities, the "wheat" available to facilities in Berthold and Bowbells is predominately durum, while the other facilities HRS wheat bushels are a greater share of total wheat bushels. The composition of the "wheat" bushels is important because, as previously discussed, durum has not exhibited as strong a market as HRS wheat for sales via unit train and shuttle vehicles. In addition, note that overlap among shuttle facility draw areas should be considered when discussing volumes, delivery patterns, and competition among markets.

Table 13. Shuttle Facilities - Draw Area Estimates for Wheat

Commodity

The draw areas defined for wheat establish the base case scenario for the remainder of the discussion regarding producer grain deliveries. The following economic analysis addresses the sensitivity of these patterns to factors such as commodity, rail rates, producer truck costs, and commercial truck delivery. The first factor, commodity effects, is illustrated by comparing shuttle-rate based draw areas for barley and corn to the shuttle-rate based wheat draw areas. The barley analysis illustrated is completed for eight of the shuttle facilities. Corn analysis is provided for a single facility, Kindred. To better manage time and form a more coherent discussion of the remaining four factors, Berthold, Jamestown, and Kindred were selected for additional individual case analysis.

Barley

Spatial comparison of draw areas for wheat and barley results in smaller draw areas for shuttle facilities in the barley market, with the exception of Bottineau. Due to production patterns and consistency among analysis (Appendix C) seven of the 10 shuttle facility locations are considered in the barley case. Facilities in Gladstone, Scranton, and Kindred were excluded. The draw areas estimated for the remaining seven shuttle facilities covered approximately 23 percent of the land area in North Dakota. Production in these draw area boundaries is estimated to be 26,354,000 bushels. Less the 119,000 bushels produced in Minnesota, the combined draw areas of the seven facilities equate to 24 percent of North Dakota's annual barley production.

Table 14. Shuttle Facilities - Draw Area Estimates for Barley

Bottineau has the draw area with the largest volume of barley -- 7.7 million bushels. Barley available in the draw areas for facilities in Berthold and Alton is over the 5 million mark, at 5.8 and 5.1 million bushels respectively. The other four facilities have draw area totals for barley that range from 3.1 to 1.6 million bushels. Considering that market discussions usually quote a 10 to 15 million bushel requirement for a feasible shuttle operation, given current market conditions, a facility would not be economically successful using barley as the sole commodity for the shuttle market. In addition, barley has exhibited some of the same tendencies as durum in its dependence on smaller freight lots, based on market comments and the UGPTI North Dakota Grain Movement Statistics discussed previously in this report.

Corn

The final commodity considered in this analysis is corn. Corn production is concentrated in the southeastern region of North Dakota (Appendix C), thus, a draw area was estimated for only the Kindred location. The draw area is estimated to include 974,000 acres. In terms of land, the draw area estimated for corn is about two-thirds the scope of the draw area estimated for wheat.

Approximately 14.2 million bushels of corn are included within the Kindred draw area boundary. Six percent of the production was attributed to Minnesota, with the balance calculated for the North Dakota portion of the draw area. This bushel estimate represents the forth largest volume attributed to a location for a single commodity. As discussed in the previous section, corn has a tendency toward adaptability to larger shipment sizes. Therefore, the shuttle rate influence is likely to be strong compared to the other commodities.

Production Densities

Production densities are an important aspect of potential adjustments in local grain delivery patterns. Individual facilities have business goals, which typically include projected grain volumes for operating profitable grain handling operations. As the projected grain volumes increase for an individual facility, that facility seeks to penetrate more distant markets and increase share for grain available in the local area. In some circumstances the additional volume may be available in the current draw areas.

When it is necessary to extend the draw area to attract additional business, the degree to which a draw area must be expanded to obtain needed additional bushels is directly related to the density of production in the draw area. For example, elevator A has a 1,000 acre draw area with an average density of 10 bushels per acre. Elevator B also has a 1,000 acre draw area, but with an average density of 30 bushels per acre. If elevator A expands its draw area acreage by 25 percent, and total available bushels increase by 2,500 bushels. This compares to an increase of 7,500 in available bushels for a 25 percent draw area acre expansion by elevator B. The previous section has included production and land area estimates for 10 shuttle facilities. A brief discussion of differences in density of production among regions and commodities completes this section. This discussion is important as it lends itself to understanding differences in the shuttle facility draw areas as they are sited in different regions of the state.

Production density, measured in bushels per acre, averaged 10.1 for the composite of the shuttle train facility draw areas. For comparison, the production densities for other areas of the U.S. Midwest region were estimated in a 1998 study of shuttle trains. Densities were estimated to be 52.4 bushels per acre in a region that covered much of central and eastern Iowa and 39.4 bushels per acre in a region covering counties in southern Minnesota, southeast South Dakota, eastern Nebraska, and western Iowa (Vachal, et. al 1998). The densities assigned to alternative Midwest locals provides an example of regional variation that should be considered in local grain market assessments.

Regarding the production densities estimated for this study, production densities vary across the state. Corn produced in the Kindred draw area raises the production density for this facility to nearly double that of Northgate, the facility ranked second when the 10 facilities are ranked by production density. Densities for Alton, Berthold, and Northgate range from 10.0 to 13.8 bushels per acre. The balance of the facilities have draw areas with an estimated production density under 10.0. Bowbells, Bottineau, and Jamestown form the mid-range group for production densities at 9.7, 8.1, and 8.0, respectively. Three elevators in the south-central and southwestern regions have the lowest production densities. The densities for the Sterling, Scranton, and Gladstone facilities average 4.8 bushels per acre. The range of production densities from a high of 24.6 to a low of 4.1 illustrates the vast difference in the landscape of North Dakota's local grain industry. It is important to recognize characteristics such as these in discussing how shuttle rates may impact the local grain industry and its infrastructure needs in specific regions.

Table 15. Production Density in Shuttle Facility Draw Areas

Rail Rates

North Dakota elevators marketed 69 percent of bushels they handled via rail between 1995 and 1999. Therefore, rail rates are a critical factor in discussing the relative competitiveness of markets and projecting future producer delivery patterns. The sensitivity of delivery patterns to adjustments in rates among elevators and railroad pricing strategies is addressed in this section. Case studies for Berthold, Jamestown, and Kindred illustrate the influence of rate adjustments on the relative competitiveness of elevators, and thus, the delivery patterns of producers.

The rail rate case study examines the effect of alternative shipping costs on the competitiveness of a local elevator. The base case elevator grain draw territories, as aforementioned, include shuttle rates for those facilities equipped to handle shuttle trains. To gauge the affect of these rates on local grain delivery patterns, grain draw territories were estimated for select case studies. In these rail rate case studies, other parameters including competing elevator rail rates and producer truck costs, are held constant while the rail rate for the selected shuttle station is replaced with a unit train rate. These results are an effective measure of the potential impact of shuttle rates on local grain delivery patterns, because in the recent past the unit train rate has been the lowest cost and most competitive alternative for shipping grain via rail.

he case study analyses show that rail rates have a substantial affect on the reach of an elevator's draw area. When the shuttle rate is replaced with the applicable unit train rate, draw areas contract in each case. The largest percentage decrease in acreage, 73 percent, is attributed to a facility in Kindred competing for corn bushels. The smallest decrease in draw area size, 23 percent, is attributed to Jamestown in the barley market. When the shuttle rate shipping option cannot be used, the Berthold and Kindred average draw areas are reduced by an average of 36 and 56 percent, respectively, for the commodities considered. Given the range of commodities and location considered in these estimates, it appears that shuttle train rates increase an elevator's draw area by approximately 50 percent, compared to the unit train draw area.

Table 16. Effect of Rail Rates on Local Grain Industry, Case Study Results of Land in Elevator Draw

As discussed, density in the draw area also is an important factor in understanding how shuttle train rates might affect local grain distribution patterns. Grain volume encompassed by draw area radii is defined by joining land area and production density values. Among the nine cases considered, the draw area volume decreases by an average 48 percent when shuttle rates are eliminated from the market. When unit train rates are applied for estimating draw area volumes, Kindred volume (considering wheat, durum, and corn) declines by 47 percent. Corn, the largest single volume commodity for the Kindred facility, decreases by 76 percent compared to the base case volume. HRS wheat volumes decline by about one-third and durum by about 16 percent.

Table 17. Effect of Rail Rates on Local Grain Industry, Case Study Results of Volume in Elevator Draw

The volume in the Jamestown draw area is 47 percent lower when the unit train rate draw area is compared to the base case shuttle draw case. Berthold experiences the smallest change in volume when individual elevator unit and shuttle train rate draw areas are compared. The Berthold facility loses access to about one-quarter of its volume it has access to with shuttle train rates for its largest volume commodity, durum, when compared to the draw area for unit train rate. Overall volume available to the Berthold facility declines by about one-third. The relative reduction in grain draw volumes are similar for HRS wheat, at 26 percent, but total volumes of HRS wheat are about one-third of durum volumes in the Berthold trade area. Barley volume, about 5.9 million bushels in the base case, is most impacted as volumes available under the unit train scenario are less than half of the volume available when the shuttle rail rate can be accessed. Land area and volume statistics are similar in conclusions. Shuttle rates provide elevators with an opportunity to penetrate a new grain draw territory. The additional territory attributed to shuttle train rates is estimated to be 18 to 76 percent farther than the most distant market available to the elevator when it utilizes unit train rates, depending on commodity and location. The average increase in volume attributed to the shuttle rate is about 50 percent.

Producer Truck Costs

The primary factor on the producer side of the farm-to-market delivery decision model is cost of delivery. The cost of delivery is based on distance and truck cost in the model applied for this study. The North Dakota producer truck fleet, as discussed previously in this report, includes an array of trucks from single-axle to semi-truck and trailer. To address the producer truck cost factor, sensitivity to truck costs is analyzed by developing truck costs at opposite ends of the truck cost spectrum - single-axle and commercial semi-truck. The results generated from this scenario are compared to results of the base case scenario - producer semi-truck costs. With this methodology, the diversity of the current producer truck fleet, changing fuel prices, opportunity for custom trucking, and potential for future investment is addressed through the range of results estimated with alternative truck costs.

The one measure of applicable producer truck costs is its effect on the size of draw areas for the case study facilities and commodities. For base case in wheat, which uses the producer semi-truck rate of $.0025 per bushel mile, Berthold, Jamestown, and Kindred have draw areas of 2,073; 2,068; and 961 thousand acres, respectively. The first of the two alternative truck costs considered in the sensitivity analysis is the single-axle truck cost. This cost is considered to be the maximum cost among possible producer truck cost scenarios. The cost for operating the single-axle truck is nearly three times higher than the semi-truck trailer, with a cost $.0091 per bushel, as detailed in a previous section. The effect of replacing the semi-truck rate with the single-axle rate is a reduction in the spans of each draw area, compared to the base case, considered in the case study analysis. The average draw area size decreases by about 44 percent. The draw area for Berthold is reduced by 18 percent, 1,704 acres, compared to reductions of 53 percent (978 acres) and 62 percent (589 acres) at Jamestown and Kindred, respectively in the base case.

Table 18. Effect of Producer Truck Costs on Local Grain Industry, Case Study Results of Land in Elevator Draw for Wheat

Considering the impact of increased producer truck costs across other commodities, Jamestown experiences the greatest reduction in size of draw area when the draw area for single-axle truck deliveries is compared to the draw area for producer semi deliveries. Considering wheat, the single-axle draw area is 74 percent smaller than the draw area defined by the producer semi scenario. The relative changes at Kindred and Jamestown are similar for corn and barley, respectively, as draw areas are reduced by about 60 percent compared to the area estimated under the producer semi scenario.

Effects vary substantially in the case of lower producer truck costs, as illustrated by the second scenario in the truck sensitivity analysis - commercial semi-truck. In this truck scenario, a producer truck cost of $.0017 per bushel reflects the costs for operations such as large farms, custom combine/trucks, elevator-owned trucks, and commercial truck operations. The commercial semi-truck rate is approximately one-third lower than the producer semi-truck operation costs due to economies of equipment use. This truck cost provides the minimum cost for the producer truck cost sensitivity analysis.

Considering wheat, the base case commodity, application of a commercial truck rate has a substantial impact on the reach of draw areas for each of the facilities. Berthold and Kindred have similar relative increases in draw area acres when commercial truck draw areas are compared to the draw areas estimated for the producer semi, enjoying more than 80 percent increase in the size of their respective draw areas. A smaller, but still notable impact, also is estimated for Jamestown at a 60 percent increase in draw area acres.

Table 19. Effect of Producer Truck Costs on Local Grain Industry, Case Study Results of Land in Elevator Draw for Barley and Corn

The impact of commercial truck rates on the draw areas for barley is less significant than for wheat. Draw area is increased by only 4 and 23 percent, respectively, for Berthold and Jamestown when draw areas are estimated using a commercial semi-truck cost compared to the draw areas estimated using the producer semi-truck cost. The commercial truck rate does provide a larger draw area for Kindred in attracting corn, compared to the draw area available for producer semi deliveries. The difference in the effect of truck rates on draw areas among facilities may be attributed to factors such as the ratio of truck-to-rail cost in the marketing equation, a relatively flat gradient in rail rates,"blanket" rail rates, and proximity of competitors. Blanket rail rates refer to application of the same rate for elevators in a region rather than scaling rates.

Table 20. Effect of Producer Truck Costs on Local Grain Industry, Case Study Results of Volume in Elevator Draw

A second measure of the relative importance of producer truck costs in the local grain delivery decision is grain volumes. The effects on draw area are similar to those found in analyzing changes in the land area included under the alternative scenarios. The average reduction in volume available to the facility when deliveries are made via single-axle truck, compared to producer semi, is 54 percent among the three facilities. Considering HRS wheat, durum, and barley, total volumes decline by 67 and 39 percent, for Jamestown and Berthold, respectively, when volume in the draw areas for single-axle producer deliveries are compared to the draw volumes estimated for producer semi deliveries. The Kindred total has a different composition as it includes HRS wheat, durum and corn. Application of the single-axle truck rates reduces volume in the Kindred draw area to less than half the volume (56 percent reduction) available in the producer semi draw area. Differences between the commodities may be attributed to factors such as differences in density across the draw area, gradient of rail rates, and commodity truck costs.

The effects of defining commercial semi-truck costs as the applicable cost for local producer grain deliveries suggest longer truck deliveries would likely occur as producer would truck farther to access lower rail rates. Considering the nine case studies (three facilities and three commodities), the volume of grain available to the facility increased approximately 55 percent when deliveries based on commercial truck costs were compared to deliveries based on producer semi-truck costs. Total volumes, considering HRS wheat, durum, and barley, increased by 13.5 and 8 million bushels (56 and 57 percent), respectively, for Berthold and Jamestown when the lower truck rate was applied. The Kindred facility had access to an additional 12 million bushels - a 51 percent increase from volumes available with the higher producer semi-truck rate. As illustrated in the array of results, local grain deliveries are influenced by many factors. The results summarized in this section provide a base for understanding the sensitivity of grain flows to inbound truck delivery costs and outbound rail rates. The range of results may be valuable in interpreting market phenomenon and forecasting needs of the local grain industry continues to seek means for competing in a global market.

Local Processor Relation Model

The second component in the discussion of local grain delivery patterns is the effect of alternative local grain market scenarios on local processors. North Dakota has long been a proponent of local processing. Its government supported the establishment of the State Mill and Elevator nearly eight decades ago. The state continues, through financial and legislative efforts, to develop local processing. Thus, understanding the influence of local grain market factors on the competitive position of the facilities is important for decision makers involved in infrastructure, economic development, and policy, as well as to producers who supply these markets. This relationship is described in the following equation:

where,

The key factor in understanding economic influence of a shuttle rate on the local processing market is in the determination of P_L, the price at the local market, where for example, the current price is a market equilibrium for supply and demand. The opportunity for the elevator to access a lower freight rate through use of a shuttle train rate effectively shifts the export market supply curve to the right. The local processor, in comparison, continues to face the same total transportation cost, but the lower cost in the competing market effectively shifts the supply curve to the left.

In the short run, the shift in the supply curve, which results from a change in the price of a competing product, will increase prices for the domestic market. Where economically feasible, the domestic market will internalize a portion of its freight rate and continue to operate. Over the long-run, competitive pressures will establish a new market equilibrium where marginal domestic buying will be eliminated and/or additional volume will enter the market.

Three case study examples illustrate how producer truck cost and elevator pricing scenarios affect economics of the local processing industry. Case studies include wheat producers close to Rugby and Forest River, and a corn producer near Colfax. The relationship is viewed through two steps. First is the determination of the elevator price, second is the producer marketing decision. The elevator price is an input to the producer marketing decision. Three alternatives are considered in the elevator pricing portion of the relationship: (1) local processor, (2) domestic or eastern market (Minneapolis, MN), and (3) western export market (PNW or Pacific Northwest). The highest return among these alternatives -- simply defined as market price less transportation cost, becomes the elevator board price among the producer options. The producer marketing decision is defined by two options, (1) elevator and (2) local processor. Producer return is defined as market price less trucking costs. It is assumed that the producer will choose that market the maximizes his net return, given the truck he operates, as this truck type will determine his truck cost per mile. The impact of alternative truck costs on this decision is illustrated through a comparison of single-axle, tandem, and semi-truck deliveries to the two markets.

In the first example of the local processor and shuttle rate relationship, the elevator maximizes its return by selling into the PNW at the shuttle rate. The highest return to the producer with any truck type is to market through the elevator for a net return of $2.22; $2.35; or $2.40 per bushel, respectively, for deliveries made via a single-axle, tandem, and semi-truck. When the shuttle option is removed from the market, the State Mill and the PNW are at par for the elevator, based on market price and freight rates. The elevator remains the best delivery option for producers, regardless of truck type. Returns to the producer are reduced by nine cents per bushel to $2.13; $2.26; and $2.31 per bushel for single-axle, tandem and semi-truck deliveries, respectively.

Table 21. Rail Rate Spreads for Local Processor vs. Shuttle Relationship: Wheat Producer - Rugby

The other two examples of this relationship are considered for producers in closer proximity to local processors. The first of these scenarios considers a wheat producer located near Forest River. The elevator is offered the same market options, the State Mill, Minneapolis, MN, and the PNW. The elevator maximizes returns with a sale to the State Mill with or without the shuttle rate shipping option. The highest return for the producer considers two market alternatives, local elevator and the State Mill. For the producer making the delivery via single-axle or tandem axle, the returns are at par for delivery to the elevator or the State Mill. When the shuttle option is removed from the market, the marketing choices and returns to the producer remain stable at $2.56; $2.61; and $2.63 per bushel. The close proximity of this elevator to the State Mill requires a more substantial rail rate adjustment for a definitive shift of traffic from the local processing truck market to a more distant rail market.

Table 22. Rail Rate Spreads for Local Processor vs. Shuttle Relationship: Wheat Producer - Forest River

The final example of the shuttle rate/local processor relation is illustrated for a corn producer near Colfax. In this market relation, the elevator is again provided with three alternatives: (1) Cargill, a local processor, (2) Minneapolis, MN, (eastern or domestic market), and (3) the PNW (western export market). Among these markets, the PNW provides the highest return to the elevator when a shuttle shipping option is available. The producer maximizes his net return, considering market price less transportation cost, by delivering to the local elevator for all

Table 23. Rail Rate Spreads for Local Processor vs. Shuttle Relationship: Corn Producer - Colfax

truck types. The return to the producer varies at $1.36; $1.41; and $1.43 per bushel for deliveries made via single-axle, tandem, and semi-truck, respectively. When the shuttle option is removed from the market, the local processor becomes the best market for the elevator. Producer returns continue to be maximized through delivery to the local elevator, but per bushel revenues are reduced by 5 cents per bushel. Thus, the impact of the shuttle rate on the ability of the local processor to compete for bushels should be a consideration in local grain delivery patterns and in viability of local processing.