April 2018

Planning and operating a duo-lateral well in Mittelplate oil field

Germany’s largest oil field, operated by DEA Deutsche Erdoel AG, sits in a national park and UNESCO World Heritage site, with development drilling restricted to one existing island.
Florian Bremer / DEA Deutsche Erdoel AG

Mittelplate field, Germany’s largest oil field, was discovered in 1966 and appraised between 1980 and 1981. It has produced since 1987, with current daily production of about 25,000 bopd. The field sits within the German Wadden Sea (Fig. 1), which has achieved status as a national park and UNESCO World Heritage site. Therefore, field development is restricted to the existing artificial drilling and production island, Mittelplate A. A plan for multilateral wells was applied successfully for the field’s development, to optimize the limited available slots.

Fig. 1. Location of Mittelplate field in the German Wadden Sea. Map modified from Wadden Sea World Heritage Map, June 2017.


Mittelplate oil field is situated in the Westholstein Jurassic trough and is related to the structural evolution of the Büsum salt dome, Fig. 2. Oil accumulations are trapped in tilted Jurassic (Dogger) and Lower Cretaceous (Wealden) formations at the western flank of the salt dome, Fig. 3. Three major regressive-transgressive cycles accumulated in the Dogger of Mittelplate, reflecting the progradation and withdrawal of deltaic sedimentation into a marine environment. Sandstone thickness and continental signature increase from the lower to the upper cycle. The lowermost mid-Jurassic reservoir unit, the Dogger beta, consists of comparatively thin sandstones, sandstone-shale intercalation and shales.


Fig. 2. A 3D view of Mittelplate field, showing the Büsum salt dome, Base Dogger beta reservoir and the fault pattern. Mittelplate development wells are shown in yellow, Dieksand ERD wells in green, and the Mittelplate-A 26 well is in red.


Fig. 3. Geological west-east cross-section of Mittelplate field.


The Dogger gamma, delta and epsilon sandstones were deposited during the following cycles of deltaic sedimentation. They are characterized by shallow marine, deltaic, fluvial and estuarine facies associations (Lippmann 2013). Marine mudstones above the Lower Cretaceous Unconformity are the major seal of the upper reservoirs. An additional intra-formational seal is formed by a thick shale sequence that separates the Dogger beta from the Dogger gamma reservoir. The oil-bearing reservoirs were sourced from organic-rich sediments of the Lias epsilon (Posidonia shale).

The Dogger beta is the most important reservoir at Mittelplate, due to the deep oil-water contact and the large areal extend of oil-bearing sandstones. The reservoir sandstones unit has a nearly constant thickness of 15 m in the central field. The reservoir consists of three sandstones (B-, C- and D-sandstone) with sufficient reservoir properties, Fig. 4. The sandstones were deposited in a delta front setting and consist of mouth bars with wedge or lentil-shaped geometry.


Fig. 4. Dogger beta reservoir correlation from central to southeastern Mittelplate field. Wells M1 and M2 are from the central portion, while the M3 and M4 near-field appraisal wells are in the southeast.


Field-wide mudstones, containing biostratigraphic markers, are present above and below the reservoir, i.e. above the Dogger beta D-sandstone and below the Dogger beta B-sandstone. The reservoir is embedded in sandstones above and below that do not belong to the pay zone. Above the reservoir, two tight silty sandstones (F- and E-sandstone) are used as markers when landing wells in the reservoir. The A-sandstone below the reservoir is a marly-to-calcareous sandstone that is tight in the central and southern field. The Dogger beta F-, E- and A-sandstones are important marker horizons for correlation and well orientation during geosteering operation, Fig. 4.


 The initial plan of the Mittelplate-A 26 well followed the concept of target points—circular reservoir areas that were allocated to a well. The well was planned as a multilateral, with two drainage points in southeastern Mittelplate field assigned to each lateral. Technical requirements comprised a tangent section above the reservoir, for the junction technology and ESP placement, as well as a kick-off point as close to the reservoir as possible. Operational strategy was to land the mainbore in the reservoir, place the junction in the nearest shale above the reservoir (between E- and D-sandstone, Fig. 4), and expose the reservoir in two horizontal liner sections with the mainbore and the lateral. Detailed planning had to consider lateral changes for reservoir thickness and properties, the structural uncertainty, and the possible conversion of producing wells into injectors.

Drilling results of exploration and near-field appraisal wells revealed a constant decline in reservoir thickness and associated reservoir properties toward the southeast, reflecting a facies trend toward more distal deposits, Fig. 4. Interpretation of the facies trend suggested a significant reduction of reservoir thickness in the drainage area of Mittelplate-A 26. Pinching out of the Dogger beta B-sandstone was known from wells to the east and south, and was very likely for the working area.

Explicit differences of the structural style of the Dogger beta over the entire Mittelplate field were shown by 3D seismic interpretation. In contrast to the central field with very little faulting, the southern part is highly structured and intensely compartmentalized, Figs. 2 and 5. North–South-trending normal faults with hanging wall displacement to the west are clearly visible on the 3D seismic at the level of Dogger beta reservoir. They indicate a vertical offset on the order of 10 m to 25 m in the working area of Mittelplate-A 26. According to seismic interpretation and well data from producing and near-field appraisal wells, the fault throw at the main faults is gradually increasing toward the east.

Production data from wells in the central and southern field support the assumption of intense compartmentalization. Pressure communication between North-South-oriented producing wells is common, whereas neighboring wells in an East-West direction do not respond to changes in pressure or production rate. Consequently, the initial drainage concept of target points was modified to a drainage of fault-bounded reservoir compartments. The intense compartmentalization, and the anticipated pressure regime, had a strong influence on well planning:

  •   Drainage area definition: three fault-bounded compartments to the east and south of producing wells.
  •   Reservoir exposure strategy: east-to-west-directed trajectory in the reservoir, to follow the dip direction of the Dogger beta and the down-faulting trend at the faults.
  •   Possible re-completion of northern wells to injectors: east-to-west-directed trajectory for best pressure support.


In a first step, two well path concepts were designed for detailed planning, Fig. 5. One of the concepts was directed in a westward direction through the reservoir, so that both laterals should expose all three compartments (No. 1 to No. 3, Fig. 5, concept blue). The alternative concept was directed in a southwestern orientation through the reservoir (Fig. 5, concept orange). Compartment exposure was assigned to different laterals of the well. The main bore should expose the eastern compartment (No. 1), while the lateral would expose the western and central compartment (No. 2 and No. 3).

Both concepts were evaluated, according to geological and technical aspects, Table 1. Important criteria for concept assessment are in the following items.


Fig. 5. Dogger beta map of the southeastern Mittelplate field, log sections of relevant offset wells and the planned trajectories for the Mittelplate-A 26 well. Faults are shown in grey; existing wells are in black; planned well trajectory concepts are in blue and orange; the final planned well path is a red dashed line; and the target area is a yellow dotted rectangle, including the drained compartments (No. 1. to No. 3.).



Wellbore length. By 2012, the longest well drilled from Mittelplate-A (not Dieksand) reached a total depth of 19,930 ft, MD. The well should not exceed 19,6854 ft, by very much, to mitigate drilling and completion risks. For the main bore, a liner length of approximately 3,280 ft was planned for the 7-in. and 5-in. liners, respectively. The 5-in. liner of the lateral was planned to reach a maximum length of 3,773 ft.

Junction placement. The most critical point of well planning was the definition of the junction point and the associated tangent section. According to plan, the 95/8-in. casing had to be drilled into the Dogger beta reservoir, and the kick-off point for the lateral placed in the shale above (Dogger beta E-shale). The limitations were the maximum length of the 95/8-in. casing, a minimum tangent section of 984 ft for ESP and junction placement, depth uncertainty of the Top Dogger beta reservoir, fault pattern and fault throw. A worst-case scenario would have been to land the well in the Dogger beta reservoir, set the 95/8-in. casing, continue drilling with the next section, and become dislocated below the reservoir without the chance to reach the oil-bearing sandstones within a reasonable distance.

Reservoir drainage. Optimized reservoir drainage depended on the fault pattern and the assumed variation of reservoir properties across the target area. All major faults in the area were supposed to be sealing, and a compartmentalized reservoir was expected. For optimum reservoir drainage, an equal drawdown and evenly distributed production of both laterals was required. Therefore, the reservoir properties of the exposed sandstones in both laterals had to be as homogeneous as possible. The anticipated reduction of reservoir thickness, and decreasing porosity and permeability toward the east, neglected the assignment of one compartment to a specific lateral. The three fault-bounded compartments had to be exposed in both laterals, as it was planned in the westward directed well path (concept blue, Fig. 5).

The optimized well path should have a sufficient distance between both laterals. Due to the geological uncertainty, a major northward shift of the junction point was not possible (Fig. 5, “final plan”). The only way to increase the north-south distance between main bore and lateral was to redesign the west turn of the main bore after setting the 7-in. liner (ahead of the first compartment bounding fault), and to extend the planned main bore to 20,013 ft, to reach the third compartment.

Reservoir engineering requirement (three compartments per lateral), westward dip direction of the Dogger beta, and the westward dipping normal faults, predetermined the east-to-west-directed reservoir exposure. Drilling in an eastward direction would have required a wellbore inclination above 90° to follow bedding. Furthermore, crossing the base of the reservoir from below, when drilling from a footwall to a hanging wall fault block, had caused severe directional drilling problems in past wells.

Intense compartmentalization associated with additional faults, not detectable on seismic, could increase the non-productive sections and hamper the pressure support for the production. Due to low seismic coverage at the southern sector of the working area, a higher number of sub-seismic faults had to be expected.

The geosteering strategy was intended to reduce unproductive sections at faults, and to keep the orientation within the reservoir. The plan was to drop sufficiently early to the lowest sandstone of the reservoir, and keep a bedding-parallel wellbore orientation when approaching a fault. By the time the well was being planned, it was unknown whether the reservoir in the target area consisted of two or three productive sandstones; a pinch-out of the Dogger beta B-sandstone seemed to be very likely. Hence, the early drop provided extra information about the stratigraphic architecture of the reservoir

Possible re-completion of producing wells to injectors: Water injection in two wells north of the drainage area of Mittelplate-A 26 would have a strong influence on the production. Depending on the well design, a water breakthrough could hamper production in an early stage. Several scenarios of well orientation, intense compartmentalization or loss of perforation length had been simulated to identify their specific impacts on the production profile. Simulation studies showed an earlier water breakthrough, a shorter production time, and a higher water cut by the end of production for the southwest-directed well trajectory.

The final planned well path was designed as a synthesis of the two concepts. The plan was a realization of 5,905-ft gross reservoir exposure in the main bore and 3,280 ft in the lateral. The reservoir exposure strategy aimed to reach all three compartments with the main bore and the lateral. According to dynamic reservoir simulation, this approach maximized the total oil production. The mainbore and the lateral should be drilled to final depths of 20,103 ft, MD, and 18,110 ft, MD, respectively, with a kick-off point for the lateral in the Dogger beta E-shale planned at 12,959 ft, MD.


Below the 30-in. conductor pipe the well was drilled in six sections, Fig. 6. Within the first three sections (20-in., 16-in., 133/8-in. casing), the well was following a planned trajectory trough the overburden until the top Dogger gamma reservoir.


Fig. 6. A geological section along the well path of the Mittelplate-A 26 main bore.


In the following 95/8-in. section, the well was drilled through the Dogger gamma reservoir and the Dogger gamma shale below, and it landed in the Dogger beta reservoir. The objective of this section was the realization of a 984-ft tangent for ESP and junction installation and landing in the uppermost Dogger beta sandstone (D-sandstone). The first objective was realized while following the plan; the second objective was done in the first geosteering operation.

As per plan, the well was drilled into the Dogger beta reservoir, and the kick-off point for the lateral was placed in the shale above (Dogger beta E-shale). The tangent section reached a total length of 1,968 ft and was ended in the uppermost reservoir sandstone, giving enough space for placing the junction in a close position above the reservoir. The wellbore inclination for the reservoir approach was chosen appropriately to level out in the uppermost sandstone, and the well could be kept for another 656 ft within the Dogger beta D-sandstone, until 13,780 ft, MD, giving an optimum position for the subsequent geosteering sections.

Reservoir exposure and perforation interval realization were done in the 7-in. and 5-in. liner sections in geosteering mode. Objective for the 7-in. liner section was to keep the well inside the Dogger beta D-sandstone and avoid an unintended reservoir exit, Fig. 7. Due to the nature of the delta front sediments, lateral changes of lithology and varying bed thickness had to be considered. No major faults were expected in this section, but sub-seismic faults seemed to be likely.

When drilling the 7-in. section, only minor changes to the well inclination had to be implemented, to follow a continuous placement in the D-sandstone. Azimuthal log data were used to model the D-sandstone thickness, showing a continuous thickness reduction from approximately 14.8 ft to 6.6 ft at the most eastern position of the well path, followed by a thickness increase to 11.5 ft by the end of the section, Fig. 7. Section TD was called at 16,850 ft, MD, with a total sandstone exposure of 3,051 ft.


Fig. 7. The geosteering objective and result of the 7-in. section in the Mittelplate-A 26 main bore.


The 5-in. liner section was the most demanding part of the well. Within this section, the well was drilled through three reservoir compartments and crossed two major faults with unknown fault throw. Intense compartmentalization associated with additional faults and changing reservoir properties, as well as pressure regime, had to be accounted for. The objective was a confirmation of the anticipated B-sandstone pinch-out, an exposure of all reservoir sandstones in all compartments, and a drop below the reservoir at the end of the section for well correlation purpose.

At the beginning of the 5-in. section, while dropping toward the base of the reservoir sandstones, the B-sandstone pinch-out was confirmed. A direct transition from C-sandstone to B-shale proved the missing B-sandstone, Fig. 8. The expected fault was encountered approximately 98 ft later than prognosis. Two faults were identified, placing the wellbore into the overlying E-shale; a combined fault throw of 66 ft was estimated by log correlation. LWD formation pressure measurements showed a significant reservoir pressure decrease between the first (60-bar pressure depletion) and the second compartment (40-bar pressure depletion).


Fig. 8. The geosteering objective and result of the 5-in. section of the Mittelplate-A 26 main bore.


A structural downward movement was maintained to move the wellbore from the D-sandstone to the underlying C-sandstone. Frequently performed formation pressure measurements confirmed the constant level of depletion in the second compartment, indicating good pressure communication within the sandstone layers. A slightly lower formation pressure, and an abrupt change of drilling performance, indicated the transition to the third compartment at a certain stage. A sandstone-sandstone contact fault, with an offset of approximately 23 ft was crossed, placing the wellbore from the C-sandstone into the overlying D-sandstone. This was recognized as the sequence from D- to A-sandstone, and it was logged in the final drop toward the base of the Dogger beta reservoir. Well TD was called at 20,046 ft, MD. A total of 6,266 ft was drilled in the 7-in. and 5-in. sections, and 5,315 ft, MD, could be attributed to the D- and C-sandstones, respectively. A net-to-gross ratio of 85% was achieved in the mainbore.

The intention of the 7-in. section of the lateral was to gain distance between the mainbore and the lateral, so a perforation was not planned. The geosteering objective of the 7-in. liner section was to navigate the wellbore from the KOP below the Dogger beta E-sandstone, toward the underlying C-sandstone. A formation parallel position in the C-sandstone should be achieved at the casing point, Fig. 9.


Fig. 9. The geosteering objective and result of the 7-in. and 5-in. section of the Mittelplate-A 26 lateral.


In the subsequent 5-in. liner section, the primary objective was to maximize exposure to the Dogger beta C- and D-sandstones in the reservoir compartments; the missing of B-sandstone appeared to be very likely in the target area. The B-sandstone pinch-out was later proven by drilling results of the lateral. The Dogger beta D-sandstone of the first compartment was extensively exposed in the 7-in. section of the main bore, therefore the exposure strategy of the lateral concentrated the geosteering efforts in the first compartment on the C-sandstone.

To compensate for the anticipated throw of the fault between compartments one and two, a drop toward the Dogger beta A-sandstone was initiated 164 ft in advance of the anticipated fault. The A-sandstone was followed until crossing the fault some 262 ft, MD, later than expected.

A vertical fault throw of approximately 49 ft was determined, placing the wellbore into the E-shale, Fig. 9. The well was dropped toward the D-sandstone and formation parallel was realigned within the D-sandstone. To acquire sufficient C-sandstone exposure in the second compartment, the well trajectory was adjusted, and C-sandstone was drilled for almost 330 ft before crossing the second major fault. The fault placed the well into the D-sandstone, which was followed until a drop was initiated approximately 330 ft before reaching the planned total depth. When TD was called at 18,110 ft, MD, a position inside the C-sandstone was attained. In the 5-in. section, approximately 2,820 ft, MD, of Dogger beta D- and C-sandstones were accumulated—a net/gross ratio of 75% was achieved in the lateral.


The Mittelplate-A 26 well was drilled horizontally through the Dogger beta reservoir. In three geosteering sections, more than 9,840 ft, MD, of reservoir sandstones were exposed. The wells set several technical records. The main bore had a total depth of 20,046 ft, MD. At that time, it was the longest well drilled from Mittelplate A and achieved the longest perforation interval (5,500 ft) of a single well. The complete perforation interval of the main bore and lateral totals 8,071 ft, a field record.

Drilling results confirmed the current geological and structural reservoir model and the anticipated fault throws at the major faults. LWD formation pressure measurements proved the concept of compartmentalization and the increasing reservoir depletion. The well was set on production in July 2015 and came onstream with an initial production rate above expectations. 


  1. Lippmann, R., “Mittelplate oil field, core description and sedimentary facies interpretation of Dogger and Wealden deposits,” RWE Dea report, 2013.
  2. This article has been adapted from EAGE paper, “History of a duolateral well in the Mittelplate oilfield Germany, from plan to perfection,” presented at the 79th EAGE Conference & Exhibition, Paris France, June 12–15, 2017.
About the Authors
Florian Bremer
DEA Deutsche Erdoel AG
Florian Bremer earned his PhD in geology from the Technical University of Berlin in 2005. Since 2007, he has worked for DEA Deutsche Erdoel AG in various departments, in exploration and production. Dr. Bremer is working as a development geologist on the German gas field, Voelkersen.
Related Articles
Connect with World Oil
Connect with World Oil, the upstream industry's most trusted source of forecast data, industry trends, and insights into operational and technological advances.