Turkey - Bati Raman Enhanced Oil Recovery Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 7916 PROJECT COMPLETION REPORT TURKEY OIL RECOVERY ENGINEERING PROJECT (LOAN S-13-TU) AND BATI RAMAN ENHANCED OIL RECOVERY FIELD DEMONSTRATION PROJECT (LOAN 1917-TU) JUNE 26, 1989 Energy Operations Division Country Department I Europe, Middle East and North Africa Regional Office This document has a restricted distribution and ma, be used bv recipients oniv in the pertormance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  TH WORLD SANK FOR OMCIAL USE ONLY Washington. DC 20433 USA O.e W D<tVwISMiMb June 26. 1989 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Turkey Oil Recovery Engineering Project (Loan S-13-TU) and Bati Raman Enhanced Oil Recovery Field Demonstration Project (Loan 1917-TU) Attached, for information, is a copy of a report entitled 'Project Completion Report on Turkey - Oil Recovery Engineering Project (Loan S-13-TU) and Bati Raman Enhanced Oil Recovery Field Demonstration Project (Loan 1917- TU)' prepared by the Europe. Middle East and North Africa Regional Office with Part II of the report contributed by the Borrower. No audit of this project has been made by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authonization.  FOR OFmFCIAL USE ONLY PROJECT COMPLETION REPORT TURKEY OIL RECOVERY ENGINEERING PROJECT (LOAN S-13-TU) and BATI RAMAN ENHANCED OIL RECOVERY FIELD DEMONSTRATION PROJECT (LOAN 1917-TU) TABLE OF CONTENTS Page No. PREFACE .......................................................i EVALUATION SUMKARY ............................................ii PART I- Project Identity .........................................1 Background ............................................... 1 3. Pro,ect Objectives and Description ....................... 2 4. Project Organization and Implementation .................. 3 5. Project Results ............................... ........... 6 6. Project Sustainability ................................... 7 7. Bank Performance ......................................... 7 8. Consulting Services ...................................... 8 9. Performance of the Beneficiary (TPAO) .................... 9 PART II Comments from the Borrower on Part I ..................... 11 PROJECT COMPLETION REPORT I. INTRODUCTION ............................................. 17 II. PROJECT IDENTIFICATION, PREPARATION ND APPRAISAL ........ 18 Origin of the Project ..... ..... ................... 18 Preparation and Appraisal ........................... 19 Project Description ................................. 21 III. PROJECT IMPLEMENTATION ................. .... ......... 22 3.1 The Bati Raman Field Carbon Dioxide Injection Project 22 Project Management .................................. 22 Project Ccst ........................................ 22 Project Disbursement ................................ 23 Procurement and Construction ........................ 24 Surface Facilities .................................. 25 Dodan Plant ......................................... 25 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  TABLZ OF CONTENTS (Cont.) Page No. The Dodan-Bati Raman Pipeline ....................... 29 Bati Raman Field .................................... 29 Project Planning .................................... 30 Well Design ......................................... 33 Project Execution .......................... ..... 34 Performance Analysis ................................ 43 Project Expansion ................................... 46 Conclusions ......................................... 48 3.2 The Raman Field Development Project .... ............. 50 3.3 The Hamitabat Gas Field Stimulation Project .. ..... 56 3.4 The Management Study ... I ............................ 58 ANNEXES 1. Categories of the Appraisals for the ."roceeds of Loan 1917-TU .................... 61 2. Categories of the Project Disbursement of Loan l917-TU 62 3. Procurement Schedule of Orders and Deliveries of Main Items of Equipment for Loan 1917-TU ................... 63 FIGURES 1. The Bati Raman Field EOR Pilot Project Application ...... 68 2. Sir.mplified Flow Diagram of the Pilot Project .... ........ 69 i. Bati Raman - CO2 Project ....... ......................... 70 4. Production History About Regions ........................ 71 5. Production & Injection History of the Fieldi .... ......... 72 6. Oil Production .......................................... 73 7. Monthly Oil Production (MSTB) ........................... 74 8. Bati Raman - CO2 Injection Efficiency .75 9. Injection History of Well 1109 .......................... 76 10. Production History of Well 1182 ......................... 77 11. Status of B. Raman Western Area Wells ................... 78 12. Produced Oil Distribution of B. Raman Western Area Wells 79 13. Pressure Distribution of B. Raman Western Area Wells .... 80 14. GOR Distribution of B. Raman Western Area Wells ......... 81 15. Production Time in "Huff & Puff' Pnplication ............ 82 16. Gas Flooding Performance of Prodo tion Wells ............ 83 17. Gas Solubility i.l Oil in Matrixes and Fractures ......... 84 18. The Gas Flooding Performance of Wells with Low Injection Rates ................. . .................. 85 19. The Bati Raman Field EOR Project Expansion ........ ..... 86 20. Cumulative Water/Cumulative Oil (MSTB) .... .............. 87  TABLE OF CONTENTS (Cont.) Page No. PART III 1. Related Bank Loans ....................................... 88 2. Project Timetable ........................................ 89 3. Loan Disbursements ....................................... 90 4. Project's Objectives and Description ..................... 90 5. Project Costs and Financing .............................. 91 6. Project Results .......................................... 93 7. Status of Covenants ...................................... 96 8. Use of Bank Resources .................................... 99 GRAPHIC Bati Raman EOR Project - Planned and Actual Disbursements 101  - i - TURKEY OIL RECOVERY ENGINEERING PROJECT (LOAN S-13-TU) AND BATI RAMAN ENHANCED OIL RECOVERY FIELD DEMONSTRATION PROJECT (LOAN 1917-TU) PROJECT COMPLETION REPORT PREFACE This is the Project Completion Report (PCR) for the Oil Recovery Engineering Project (Loan S-13-TU) and the Bati Raman Enhanced Oil Recovery Field Demonstration in Turkey, for which Loans S-13-TU and 1917-TU in the amounts of US$2.5 million and US$55.2 million, respectively, were approved on November 30, 1978 (Loan S-13-TU), and on November 24, 1980 (Loan 1917-TU). The lu,ans were closed on December 31, 1987. Loan 1917-TU was not fully disbursed and the last disbursement was on July 18, 1988. The PCR was jointly prepared by the Energy Operations Division, ,..ountry Department I of the Europe, Middle East and North Africa Regional Office (Preface, Evaluation Summary, Parts I and III) and the Borrower, (Part ii). Preparation of this PCR is based, inter alia, on the Staff Appraisal Report; the Loan, Guarantee, and Project Agreements; supervision reports; correspondence between the Bank and the Borrower; and internal Bank memoranda. _ 1  - {i - PROJECT COMPLETION REPORT TURKEY OIL RECOVERY ENGINEERING LOAN (S-13-TU) AND BATI RAMAN ZNHANCED OIL RECOVERY FIELD DEMONSTRATION PROJECT (LOAN 1917-TiJ) EVALUATION SUMMARY Introduction 1. The project was the second Bank operation in Turkey's petroleum subsector and the first Bank financed project anywhere to utilize enhLanced oil recovery (EOR) technology to increase oil production. A study financed under a preceding engineering loan (S-13-TU) for US$2.5 million had recommended carbon dioxide (C02) injection into the oil reservoir and this project was designed as a follow up pilot project (PCR, Part I, paras. 2.02-2.03). The Bank loan of US$62.0 million, approved on November 18, 1980, was given to Turkiye Petrolleri A.0. (TPAO), the national oil company, with the guarantee of the Republic of Turkey. Obiectives 2. The main objectives of the project were to expand Turkey's recoverable reserve base through: (a) application of EOR technology at Bati Raman; (b) development of the newly discovered oil reserves in the Raman oil field; and (c) reserve evaluation and stimulation of gas wells in the Hamitabat field to increase deliverability. ImDlementation ExDerience 3. Although delayed by three years due mainly to technic I and procurement problems, as well as change in scope of EOR method, '- project was completed with savings of US$24 million (23.4%) in total proje.. -b-s. The savings can be attributed in part to prudent use of local contractors for the civil works portion as well as to the devaluation of the Turkish Lire vis-a-vis the US dollar (PCR, Part III, Section 4). The disbursement of the loan was slower than forecast reflecting implementation delays experienced by TPAO (PCR, Part I, para 4.02(a) and para 9.01). The unutilized portion of the loan (US$7.4 million) was cancelled retroactively to March 18, 1988. The performance of the consultants and contractors was rather uneven, particularly in the early stages of the EOR component when TPAO apparently did not closely supervise its consultant. This situation improved in the course of implementation, when adequate coordinating arrangements were put in place to ensure satisfactory performance (PCR, Part I, para 8.01). The Bank's supervision was satisfartory and a reasonable continuity of staff working on the project was maintained throughout implementation.  - iii - Biua 4. While the EOR method initially identified as having the most promise ('huff n puff' method) was changed to gas line drive, the physical objectives were saiccessfully met. Oil production from Bati Raman heavy oil field increased from 900 barrels per day (B/D) to 6,600 B/D by 1989 (PCR, Part ', para 5.01). Gas deliverability from the Hamitabat aas field was increased fcurfold from 15 million cu.ft. per day to 60 million cu.ft. per day in 1989 as a result of the well stimulation program. For the Raman field, TPAO was able to recommence oii production whereby the net yearly oil production is now over 200,000 barrels. Sustainabilitv 5. TPAO's oil production from Bati Raman is expected to almost double by 1990 as the positive results of the pilot test project are being soplied to other areas of the field. The C02 injection EOR process would continue to yield net economic benefits at an acceptable level even in a lower international crude price scenario. Findings and Lessor.E 6. The rather long delay in conpleting the project components has been attributed to various factors including changes in project design necessitated by new developments during implem,ntation (PCR, Part I, para 9.01); however, a salient lesson of the project experience points to the need for ensuring that all tasks, whether to be performed by the implementing agency, contractors or conaultants are coordinated and strictly monitored if unexpected delays are to be minimized.  Part I 1. Pro1ect Identity Project Name: Bati Raman Enhanced Oil Recovery Field Demonstration Project Loan No: 19i7-TU RVP Unit: Europe, Middle East and North Africa Region Country: Turkey Sector: Energy Subsector: Petroleum 2. Background 2.01 The project had its genesis In the heavy burden Turkey's oil imports placed on its balance of payments. The rapid industrial -'xpansion of the Turkish economy since the mid 1970s and the slowly declining domestic oil production resulted in an increased dependrtce on imported oil and petroleum products. While petroleum imports absorbed 60% of merchandise export earnings in 1978, they absorbed all of the merchandise earnings in 1979 and were expected to surpass the merchandise earnings in 1980. In this situation, the energy policy of the Government of Turkey (GOT) focused on restraining oil imports by developing indigenous energy resources such as coal, lignite and hydropower, by implementing measures to improve the efticiency of energy use and by augmenting indigenous oil and gas production. 2.02 The Petroleum Exploration Project, covered by Loan 1916-TU (approved and signed on the same dates as the subject Project), was completed by he end of 1985, i.e. about two years earlier than the subject Project. The PCR on that Project has concluded that the Project was a qualified success, inasmuch as it strengthened TPAO as an institution and added about 14 million barrels of oil to TPAO's established recoverable reserves, although the exploration effort was not a commercial success. Under the Project, TPAO's oil importing and refinery activities were separated from its exploration activities on the argument that TPAO's financial resources should be linked to the success of its exploration program. However, the outcome of the restructuring has been different from that originally envisaged at appraisal, and the dividends from TPAO's refineries and marketing subsidiaries are now financing TPAO's exploration effort because of the fall in crude oil prices and GOT's pricing strategy for petroleum products. 2.03 In early 1978, TPAO requested Bank assistance in developing a program to increase the ultimate recovery of the Bati Raman oil field, the largest oil field in Turkey, but producing heavy oil, through the use of enhanced oil recovery (EOR) technology. U3ing financing provided by an engineering loan (S-13-TU) of 1978 for $2.5 million, a group of consultants undertook a comparative evaluation of alternative EOR technologies applicable to heavy oil to determine the optimal enhanced recover-, method. This study indicated that carbon dioxide (C02) injectioji into the oil reservoir should be the recommended EOR method, a conclusion that confirmed the results of TPAO's own studies and was further confirmed by two other independent consultants. The Bati Raman Component (Loan 1917-TU) was thus a pilot project to test a system based on the conclusions of the engineering loan. Since the engineering loan was refinanced under the subject loan, this PCR should be considered as applicable to both the engineering loan and Ln 1917-TU.  -2- 2.04 During the final stage of the EOR study financed under the engineering loan, GOT also requested Bank assistance for: (a) the development of the newly discovered extension of the established Raman oil field; and (b) the assessment of the production potential of the Hamitabat gas field in Thrace in northwestern Turkey. 2.05 Part II of the PCR, prepared by TPAO, although a good report assessing the problems faced in implementation from a technical perspective, has serious limitations in addressing key implementation issues such as procurement, performance of consultants, contractors, the Bank and TPAO, and an assessm'ent of the Project results vis-a-vis appraisal expectations. 3. ProAect's Objectives and Description 3.01 The project has three main objectives: (a) to increase Turkey's crude oil production in the short term by developing the newly discovered oil reserves in the Raman oil field; (b) to expand Turkey's recoverable reserves base through the application of new euihanced oil recovery techniques and to enhance its medium-term petroleum production capacity; and (c) to evaluate the gas reserves and production potential of a newly discovered gas field. 3.02 The project comprised the following four components: (a) The Bati Faman Field Test - Following completion of the comparative evaluation study, which recommended the application of CO2 technology, it was proposed to undertake a demonstration test to determine whether the process would in fact produce the expected results. If successful, the results of this two-year demonstration test would provide TPAO with the basic operating data required for it to design an efficient full field development project. The field demonstration test of the chosen carbon dioxide EOR technology would be carried out on 10 percent of the reservoir. It would require the drilling of five wells at the Dodan gas field to produce carbon dioxide gas, the removal of sulphur fron the gas, the laying of a 75 km pipeline to transport 55 million cubic feet per day of carbon dioxide gas from the Dodan gas field to the Bati Raman oil field at a pressure of 2000 psi, the drilling of an additional 5 wells in Bati Raman, the preparation of a total 30 wells to handle C02 injection and oil production, and the installation of associated equipment in Dodan and Bati Raman. Workover rigs, a cementing truck, specialized laboratory equipment, and equipment and seIeices comprising technical assistance for testing, monitoring and evaluating the results, were also included; (b) The Raman Field Development - This component would assist in the expansion of production from tne newly discovered northern extension of the Raman oil field, through the drilling and completion of 18 new production wells and the installation of associated surface and subsurface equipment. In addition, technical assistance for a reservoir study was proposed in order to determine the optimum approach to a secondary recovery program for the entire field;  - 3 - ( ) Thrace (HAmitabat) GaR Fleld - The production potential of the Hamitabat gas field would be evaluated by fracturing and testing six to ten wells to provide reliable estimates of gas reserves and field productivity and to determine the optimum utilization of its gas. Although seven wells had been drilled and thre' were already producing a total of five million cu.ft. per day of gas, neither tht reserve estimates nor the production fc ecasts lent sufficient confidence to permit the implemeitation of projects designed to consume the gas over an extended period of time. Technical assistance for a atudy of gas reserves and optimum utilization of the gas was also included; and (d) Technical Assistance - To strengthen TPAO's management and organization, TPAO would be provided with specialized assistance for improving its operational productivity aiid its management organization. At appraisal, TPAO was running a huge operation with little of the work contracted out. The organizational structure, systems of internal communication, control of operations, warehousing, accounting and financial management, financial planning systems and managemtilt of its subsidiaries would be carefully reviewed with a view to improving TPAO's performance within set policy objectives. 4. Prolect Organization and Implementation 4.01 Proiect Organization. Since the Bati Raman Field Demonstration Test was an inherently risky project, work on the project had to proceed cautiously. Accordingly, the following steps were taken to ensure the success of this and the other project components: (a) TPAO would create, for each project component, a special project implementation group with qualified and experienced staff. The organization and ctaffing of the groups, their technical responsibilities and financial authoricy were agreed upon with the Bank before project implementation. The effectiveness of the groups and their staffing were to be constantly reviewed during project implementation and if the capabilities of these groups fell below acceptable levels by the loss of qualified technical personnel, TPAO agreed to secure adequate outside technical assistance; (b) Although TPAO had overall responsibility for project implementation, it would call upon foreign expertise in areas where it lacked technical capabilities. Accordingly, for the Bati Raman EOR Field Demonstration Test, it engaged separate, experienced consultants for detailed engineering design, for supervision and start-up and for monitoring and evaluating the field tests. For the Raman field, TPAO would engage consultants to carry out a reservoir study to help determine the optimal EOR program for the field. For the Thrace gas field, TPAO would engage specialized service companies for fracturing and testing of the gas wells to help evaluate the production potential and to provide reliable estimates of its gas reserves. TPAO would also engage consultants to undertake a study of the technically and economically optimuri utilization of tne gas reserves, once  the gas field's production potential was established. Well logging operations for the Raman, Jodan and Hamitabat fields and for exploration wells would be provided I/ by a competent foreiSn contractor; (c) As the carbon dioxide injection enhanced recovery method was being applied to heavy oil for the firs* time under the conditions obtaining at Bati Raman, TPAO was required to implement this subproject in stages. TPAO was required to submit a phased implementatio2 program before October 1981 for the Bank's approval, to ensure that each phase was satisfactorily implemented and its progress r viewed before work was exteD4ed to the entire 30-well program of the subproJect. Furthermore, TPAO alao agreed not to undertake field tests of the C02 injection in the central section of the reservoir until the Bank and TPAO agreed that the results of the tests justified extending the tests to the central section; and (d) To benefit from the experience of foreign companies and agencies, TPAO agreed to establish by October 1, 1981 on advisory panel (independent of its consultants) consistivg of at least three experts experienced in heavy oil recovery or C02 injection to review the progress of the tests and resolve unforeseen problems. 4.02 Implementation. The details of implementation of the various project components are given below: ka) Bati Raman EOR Field Demonstration Test. This component was completed over three years behind schedule due to slippages that could be attributed to staff shortages, delays in procurement and in the inadequate performance of consultants and contractors. The detailed reasons for the delay were as follows: (i) Decreasing real staff salaries caused serious staff morale probiems in 1980 and 1981 and a large exodus of technical and managerial staff. The shortage of technical personnel and changes in the production group management adversely affected project implementation; r (ii) The poor performance at least initially of the expatriate consultant charged with the design, procurement and supervision of the project and the sluggish performance of the Turkish consultant entrusted with the design work for civil construction; 1/ Para 4.01(b) - Although this was the understanding of the appraisal mission, it was not explicitly stated in the Project kgreement (see SAR, para. 5.23).  (iii) Delays in the procurement of equipment and services due to the large number of contracts (over 400) that had to be reviewed and cleared by the Bank as well as the inexperience and lack of familiarity of TPAO staff with the Bank's procurement procedures. This caused delays all around in implementation of various components such as drilling and work over of Bati Raman and Dodan wells, construction of surface facilities, construction of gas pipeline from Dodan _o Bati Raman, etc.; (iv) Poor coordination by TPAO of the work of the various entities involved in project implementation such as the consultants and contractors, resulting in delays in decision-making and contract awardsl'; (v) Faulty design of the unloading valve and instrumentation of the C02 compressors and of the pumping unit for the Selexol plant leading to a four-year delay in commencing the C02 EOR pilot test; (vi) Change in the design of C02 gas injection method from 'huff n puff' to gas line drive in view of the discouraging preliminary results and the occurrence of C02 break- through in several wells which was not foreseen in the consultants' design; and (vii) Delay in the installation of surface facilities by Turkish contractors at the Dodan plant Ultimately, C02 injection in the project area cummenced in March 1986. However, certain technical problems such as the unsatisfactory functioning of the control panels of the compressors arose and the plant was shut down on May 1, 1986. The problems were corrected and steady C02 injection at 15-20 million cft. per day started in September 1986. Corrosion inhibitor has been injected in che wells and TPAO is monitoring possible corrosion in C02 wells as well as in surface facilities on a regular basis. One continuing problem is the excessive Selexol consumption in the C02 processing plant (over six times the designed consumption) which is raising operating costs. Efforts are under way to resolve this problem. In the interim, TPAO has procured C02 conservation facilities such as recycling compressors to reinject th.' C02 produced with the oil at Bati Raman. l/ Para 4.02 (a)(iv) TPAO disputes this point and claims that it has alv.~,ys had a good coordination with consultants and contractors. This paragraph reflects the view of the Bank staff.  (b) Raman Field DeveloDment. This component has been satisfactorily completed but is over two years late. Initially, the component was commenced using TPAO's own materials but the stock was replenished when the Bank-financed materials began arriving at TPAO's stores. In mid-1986, about two years later than the appraisal estimate, the expatriate engineering consultants have submitted satisfactory reservoir studies on the Raman field, as well as the nearby Garzan field, which was added to the project scope. Based on the consultant's recommenaations, infill wells have been drilled and additional d-ta is being collected prior to undertaking a pilot water flood at the Raman field; (c) Hamitabat Gas Fleld. Fracturing operations, expanded from the original six to twenty-four wells, were conducted with the help of a renowned expatriate firm. These were completed ! November 1984 and have been very successful. Well productivit. increased five-fold as a result of fracturing; (d) Management Studv. The study, conducted by an expatriate firm of consultants, was completed in April 1987. TPAO has reviewed the study and is implementing some of the recommendations; in particular, TPAO is taking steps to insta'll an MIS system. TPAO's staff have undergone training with the consultants to enable them to efficiently run the MIS operations. 5. ProJect Results 5.01 Except for a reduction in the number of wells in the Raman component from 18 co 17 and an increase in the stimulatiLn program in the Hamitabat gas field from 6 to 24 wells, the Project was comoleted as planned at appraisal, though over three years late. It also substantially achieved its physical objectives. The Project demonstrated che successful application of the C02 gas injection EOR technology in the pilot area. T'e 'huff n puff' method initially recommended by the consultants did not pruve effective, but the change to the gas line drive method yielded positive results. The total oil production in the pilot area and in the surrounding area, which was affected by the CO2 injection, increased from 300 B/D before the start of the C02 injection to 5,600 B/D (April 1989). Overall, the oil production in the entire Bati Raman field increased from 900 B/D just prior to the C02 injection to 6,600 B/D (April 1989), 90% of this increase being attributable to the EOR technology applied. In view of the successful outcome of the Project, TPAO is planning a cautious C02 injection program to increase production to 10,000 B/D by 1990, an output achievable without major capital investments. 5.02 Raman and Garzan Fields. The reservoir studies for the Raman and Garzan fields have been satisfactorily completed. They have evaluated the reservoir potential and made recommendations for increasing production. Practical results will follow from implementation of the recommended pilot tests, which are expected to comrmence ir 1989, and the development strategy based on the experience gained by TPAO.  5.03 Hamitabat Gas Flield. TPAO expanded trie stimulation program for their wells and has improved gas proiuctivity from the tight sands. Prior to the well stimulation program (hydraulic fracturing), the Hamitabat field was producing about 15 million cu.ft. per day. As a result of the reservoir fracturing operations financed under the project, the Hamitabat gas field production has been ir.creased about fourfold to 60 million cu.ft. (April 1919) per day. However, currently, the field is producing at only half its capacity due to the fact that the nearby Ambarli Power Plant, has been temporarii) shifted to Russian gas, secured under a take-or-pay arrangement because of the absence of infrastructure for the transmission and distribution of gas to other consumers. This situation will change when, the development of the IstalOul gas distribution network and other similar dis.tribution networks is completed, at which time the Russian gas will mostly be absorbed and the Ambarli Plant switched back to Hamitabat gas. 5.04 Management Study. The stuldy, conducted by expatriate and Turkish consultants, identified various areas for improving the organization and management of TPAO. Steps to implement a plan , install an MIS system in TPAO are under way. 6. Proiect Sustainability 6.01 TPAO has been able to complete the project with cost savings of about 23% compared to the appraisal estimates, despite the change in project scope and protracted implementation delays. In spite of the initial setback with the cyclIc C02 gas injection (huff and puff) process envisaged at appraisal, TPAO managed to alter the C02 gas injection to line drive with significant success. At the current level of production of about 6,600 bbl/day (2.4 million barrels/year), exiEcted to increase to 10,000 bbl/day (over 3.5 million bbl/year) without significant capital investment in C02 injection facilities, the Project is economically profitable and well sustainable in the future. The net benefits would continue at the present acceptable level, except in the event of a collapse of oil prices. 6.02 TPAO is currently producing Bati Raman crude at a proe ztion cost of about US$4/bbl. With the conservation of the C02 produced with oil, and recycling it into the Bati Raman reservoir and increasing field productior. to 10,000 bbl/day by 1990, the production costs are expected to be further reduced by about 20%. It is, therefore, considered that the C02 EOR process at Bati Raman would continue to yield net economic benefits at an acceptable level even in a lower international crude price scenario, and would contribute toward improving the balance of payment problems facing the country. 7. Bank Performance 7.01 The Bank performed well under the Project. It proceeded with caution in handling a technology that had not been proven in the conditions obtaining in Bati Raman. It counselled TPAO appropriately as problems arose in the course of project implementation and while doing so, it maintained good relations with GOT and TPAO. For example, in December 1984, when the production had sharply declined to 250 barrels/day, the Bank advised TPAO to run a complete reservoir pressure survey in the pilot area and to do some formation logging as the data would provide useful inputs in later C02 pilot monitoring and modelling. The Bank further advised TPAO to seek assistance from international oil companies at the critical time of supervision of the CO2 injection and monitoring of reservoir behavior. In the case of the gas wells at Hamitabat, the Bank suggested that long duration  - 8 - isochronal tests in some key wells be carried out since TPAO had only tested the wellc for very short periods. All these suggestions were accepted by TPAO. Close liaison by Bank staff with TPAO ted the consultants in reservoir performance helped TPAO in making timely changes in project execution. A particularly useful contribution was made by Bank staff in their discussions on critical occasions with TPAO, the consultants and suppliers (e.g. of C02 compressors), when a near stalemate among them threatened serious delays in project execution. 7.02 Perhaps in two areas, the Bank's performance might have been better. Knowing that this was among the Bank's earliest involvements in the petroleum subsector and that TPAO had no experience of the Bank's procurement procedures, greater attention could have been paid to educating TPAO's staff in these procedures and to better packaging of the goods and services procured under the Project. The Bank could also have conducted a more rigorous scrutiny of TPAO's contracts to ensure a tighter control of the performance of consultants and contractors. The Bank's review and questioning did result in better terms for TPAO in the consultants' contract in the Bati Raman component. 7.03 With regard to supervision too, the Bank's performance was very satisfactory. The Project was supervised at regular intervals except for the last one in December 1987, which was mounted 17 months after the previous one in July 1986, arising from some staff discontinuity during the reorganization. Otherwise, there was also reasonabl' continuity of Bank staff working on the Pro2ect. 8. Consulting Services 8.01 From mid-1981, TPAO engaged an expatriate firm (later assisted by a Turkish firm) as consultants to help in detailed engineering design, materials procurement, on-site supervision of surface facilities and monitoring of the CO2 pilot test. The performance of the consultants was rather uneven. For example, there were frequent changes in the design of the portable heaters, one of the Selexol towers and the gas separators as correctly alleged by the Turkish fabricator. In the early stages, there was inadequate follow-lip of procurement to ensure that suppliers met specifications on some equipment. There was also some excessivre billing and overcharging for the engineering services which were raised with the firm at the insistence of the Bank. The excessive billing was finally corrected and it was agreed as a corrective measure that TPAO would obtain from the consultant monthly budgets of manhours and jobs in full detail. For all these reasons, the Bank did consider the consultant's performance in 1983 as unsatisfactory. Ir 1984, their contract was renewed as TPAO was reluctant to have a change at that stage. With better supervision by TPAO and the active intervention of the Bank at critical stages, the consultant's performance improved in due course and could be considered satisfactory. 8.02 The performance of other consultants such as those under the Raman and Garzan component and the expatriate ,irm of management consultants was, generally speaking, satisfactory.  - 9- 9. Performance of the Beneficiary (TPAO) 9.01 The Project was delayed by over three year. mainly because of staffing problems, r:ocurement delays, and the unsatisfactory perfcrmance of consultants and contractors. The delay was also partly due to changes in project design necessitated by new developments during implementation. The delay, for which TPAO was to some extent responsible, could have been largely avoided with proper coordination of all tasks and a strict monitoring and enforcement of contractual performance both by contractors and consultants.l/ Part of the delay in procurement was no doubt due to TPAO's unfamiliarity with the Bank's procurement procedures, but it was also partly due to TPAO's own internal problems. At the same time, TPAO deserves credit for successfully completing all the components of the project and for its responsiveness to reasonable suggestions by the Bank. 1/ TPAO does not accept the view that the change in injection pattern contributed significantly to project delays. Again this represents the opinion of t'ie Bank staff.  - 10 - PART II (PCR PREPARED BY TPAO) AND COMMENTS FROM THE BORROWER ON PART I  - 11 - COMMENTS FROM THE BORR(CER 'N 'AP: T Page 1 of 5 U _ _ _ _ ___ e - ! hal i Pt I P | I i rt 4 I, %It1il 111 . - . .... .e eee - A.. .. d 2 P - ,s'.i-* r' ~~~~~* - r, M- *2 42 .. . . t ; ~~'ar T8*...  - 12 - Page 2 of 5 of foreign companies particularly at Middle East area, however this has not affected the implementation of the Bati Raman project. Because, the project has been conducted almost by the same managerial staff from the beginning up today. The poor performance of the Item 4.02 la) (iii): expatriate consultant alreadv mentioned. No need to amend Although the lack of familiarity of TPA0 staff with the Part I. Bank's regulations might have contributed to delays in the procurement and construction phases of the project, the main reason is attributable to slow processing in initial review of RFQ ducuments, evaluation reports and subsequent contract documents, and the concurrance of Letter of Credits by the World Bank. Another fact caused a del.ay in the proJect implementation has occurred as follows. During the time elapsed between the finc.Iization of S-13 T" and the date Loan 1917 TU became effective, the expatriate consultant temporarily suspended their engir.eering activities due to unpaid invoices which also res.uted s'hut 6-month delay in the mnplementatiorn of the project. These comments given as a Itern 4.0. (a) (i;): footnote under para 4.02 (a) (IV) PA' al'wa.s F.ave had a qw2:d ^.-r Inatian Wi . nunsultants and :-ntractors whic-. we Lon t oe.len'e r,s -:i ad., lay .n lecision-making nd czrntract awarls dar; the ;rr..ect irmple-entation. in c.rar , T ' s feref - by means of this ccnri:nato:n a4ain.a :.r :ed iemanrs -aricas entities :.y-::ei_ ; te The Bank disagrees as the change t 4.; a) a. in gas injection did result in implemeatation delays. .anqe in t-. gas ln'ecrt-o-n :oar-n fr:n !.nf- anrd pff to qjas lIne drIve os n at s eas n a uais.- idoIa; sn the prcect impleme-.tatl:.n, r.t a cor.se-. enoe -f - t-: .;eld oi1ot an,loed in the pro,ect area. Infact, T'A's ale:r..ss on the onteruretation of field results and de-ision mn cn.'ern : t,o ,as 'line drove has resulted in today's success-.l ';eli aopulLcation.  - 13 - Page 3 of 5 Para 4.02 (a)(VIII) has been Item 4.02 (a) (viii) deleted. Since it is related with the Hamitabat gas field, it should be omitted from the section related with the Bati Raman field. Para 5.01 has been suitably Item 5.01: amended, Starting from 6th ltne, this paragraph may be revized as follows consisting of latest production figures: 'The "huff and puff' method initially recommended by the consultants did not prove effective, however field results showed that most of the production increase was due to the gas drive mechanism in the reservoir and therefore the injection pattern was changed from "huff and puff" to gas line drive which yielded positive results. The total oil production in the pilot and surrounding area (CO2 affective area) increased from 300 B/D before start of the CO2 injection to 5600 B/O in April 1989. Overall, the oil production in the entire Bati Raman field increased over 6600 B/I in April 1989 whereas it was estimated to be around 900 B/D on primary production, 90 % of this increase being attributable to the EOR technology applied. In view of the successfull outcome of the Project, TPAO is planning a cautious CO2 injection program to increase production to 10 000 B/D by 1990, an output achievable without major oapital investments.' Para 5.03 has been amended. Item 5.03 Prior to well stimulation program (hydraulic fracturing), the Hamitabat fieid was producing about 15 million cu.ft.,day. As a result of the fracturing operations financed under the pro;ect, the Hamitabat gas field production has been increased about 4 fold to 60 million cu.ft.!day. . / . .~~~~~~  - 14 - Page 4 of 5 Para 6.01 has been amended. Item 6.01: %he current level of Bat. Raman field production of about 6600 B/D may be ins. rted into this paragraph. Para 6.02 has been amended. Item 6.02: In our records, the operating cost of currently producing Bati Raman crude is about 4 USD/Bbl. Para 7.01 has been suitably Item 7.01: amended. The production of Dati Raman field had never declined to 250 8/D in his history which can be seen in Figure 6 of Part II. The production in the pilot area was about this range due to the low reservoir pressure before start of the gas injection. No change needed in Part I Reservoir pressure sarvey has been carried out in the field of the FCR. as a routine operation since the beginning of the field production. Formation logging has also been applied to all Bati Raman wells as a usual procedure in order to provide useful data for the evaluation and better understanding the reservoir conditions. No change needed in Part I The reliability of deliverability tests does not depend of the PCR. on the isochronal time selected, but the extended flow time is important. This fact has been considered by TPAO in carrying out the deliverability tests in Hamitabat wells, and the extended flow periods were long enough to reach tle semi-steady state conditions. These comments given as a Item 9.01: footnote under Para 9.01. As mentioned above (Item 4.02 a-vi) changes in the injection pattern cannot be considered contributory to delays. Also, our other comments on delays must be taken into account for this paragraph. .1/. .  - 15 - Page 5 of 5 For Part III of the Project Completion Report, we agree with the text and do not have ary comment on that. Ai,o, please note that TPAO's Project Completion Report dated November 1, 1988 can be used as Part II of the final PCR. We hope our above comments will be helpfull in the revision of your final report. Sincerely Yours, TURKISH PETROLEUM CORPORATION '"~~ t^-rAlk OK/CS 18.5.1989 v 'UZ A..  - 16 - TURKEY PROJECT COMPLETION REPORT THE BATI RAMAN ENHANCED OIL RECOVERY FIELD DEMOiNSTRATION PROJECT (LOAN 1917-.JU) Turkish Petroleum Corporation Production Department  - 17 - TURKEY PROJECT COMPLETION REPORT THE BATI RAMAN ENHANCED OIL RECOVERY FIELD DEMONSTRATION PROJECT (LOAN 1917 - TU) I. INTRODUCTION 1.01 Like many countries in the world, petroleum supply has become a focal point of the economic pioblems that Turkey faces today. The share of crude oil in Turkey's overall energy requirements has increased sharply over the last two decades from 20 percent in 1960 to 45 percent in 1986. Although Turkey was producing approximately 3 million tons of crude oil per year, the rapid industrial expansion of the economy since the mid-1970s resulted in an increased dependence on imported oil and its products. Domestic oil production, which declined slowly through most of the 1970s, currently accounts for about 14 percent of Turkey's overall oil requirements of about 16 million tons. 1.02 The Government of Turkey (GOT) has made concerted effords to reduce the growth of oil consumption by replacing it with lignite and hydropower for electricity generation, and with coal for process heat in industries. While these efforts have helped to stem the growth of oil consumption, the potential for additional petroleum- substitution projects was limited. Therefore, a new strategy was needed to be established on Turkey's overall energy requirements. The efforts to augment domestic pe.roleum production through the introduction of enhanced oil recovery (EOR) technology for older and most,y heavy oil fields, and through intensified exploration and more rapid development of newer oil fields are the important  - 18 - facts of this strategy. To increase the indigenous oil production, GOT encouraged The Turkish Petroleum Corporation (TPAO) to conduct research ON EOR technology and implement EOR methods when feasible in the fields as pilot application. It is in this setting that the bati Raman EOR field demonstration project, and the Raman and Hamitabat fields development projects were conceived. These projects were the Bank's first involvement in the production of hydrocarbons in Turkey. II. PROJECT IDENTIFICATION, PREPARATION AND APPRAISAL Origin of the Project 2.01 TPAO owns and operates 30 oil fields, of which about 12 have very small recoverable reserves of about 700 000 Bbl, or less, and 5 gas fields. The oil in place of TPAO's oil fields -s calculated as 3.810 billion STB in total of which 80 percent is under 20 AP! and considered heavy oil. The primary recovery factors uf these reservoirs vary between 1 percent and 40 percenit depending on the producing mechanism, oil and rock properties. 2.02 The Raman and Bati Ramnan fields are the most important fields of TPAO. Bati Ranara contributed only 12 percent of annual production, although it is the largest oil field so far discovered in Turkey with an estimated initial reserve of about 1.85 billion STB of oil. The recovery factor by primary techniques for the Bati Raman field is estimated at between 1.5 and 2.0 percent of the original oil in place. This low recovery is due to the unfavorable properties of the oil (such as low gravity, low solution gas, high viscosity) and the absence of any appreciable driving mechanism. Because of these adverse reservoir and fluid characteristics, the field showed a fast production decline since the discovery date of 1961. The original pressure of 1800 psi dropped to about 400 psi and the daily oil production which was close to 10000 BPD in maximum in 1969 dropped to about 1500 BPD in 1985. This rapid production decline of the Bats Raman field led TPAO to consider intervening with en'lanced oil recovery (EOR) techniques.  - 19 - 2.03 EOR Technology in the world has increased since 1976. rhe primary stimulus for this activity was a significant rise in oil prices. Although oil prices peaked in 1981 and have since declined, extensive progresses in research and field applications over this period have resulted in a better understanding of the fundementals of processes applied. Since conventional primary and secondary methods are expected to recover only about one-third of the oil originally in place, the remaining two-thirds of the oil reserves,which are nonproducible due to adverse fluid properties and reservoir conditions,constitute the main target for EOR applicaticns in the world today. Therefore, the resources to which EOR methods may economically be applied have a great importance in the oil industry. Preparation and Appraisal 2.04 In view of above considerations, in early 1978, TPAO requested IBRD assistance in developing a program to increase the ultimate recovery of fihe Bati Raman oil field through the use of EOR technolr7y. Because a number of methods which may enhance the oil reco ry could principally be used, it was therefore decided to proceed in 3 stages. '. To carry out a comparative evaluation of different EOR methods to determine the most appropriate method to be used. 2. To apply a pilot test of the method chosen to evaluate its applicability to the field, and, 3. If positive results are obtained by the pilot test, to extend the application of the EOR method used to the whole Bati Raman field. 2.05 An engineering loan (S-13 TU) for USD 2.5 million for the first stage of the project was approved by the IBRD's Executive Directors in November 1978.  - 20 - 2.06 Using the funds provided for realization of the first staye, TPAO contracted a consultant group (Intercomp and Godsey - Earlougher) to carry out an engineering study to determine the optimal techniques for increasing the oil recovery from the field. In 1980, a comparative engineeting study of alternative EOR schemes to be applied in the Bat. Raman project was completed by the consultant. In this study, 6 EOR schemes were considered. 1. Improved recovery bI' chemicals, 2. In-situ combustion (dry and wet), 3. Steamflooding, 4. Steamflooding using USSR mining technology, 5. Waterflooding, 6. Carbon dioxide or Dodan gas application. 2.07 In this comprehensive prefeasibility study, the reservoir response was evaluated providing project cost analysis for various EOR schemes. The study concluded that both carbon dioxide and steam injection applications were favorable. However, immiscible carbon dioxide application appeared more feasible because of the nearby Dodan carbon dioxide gas reserves and the high initial investment cost of steam injection. As is a standard practice for EOR projects, a de-nonstration test should be undertaken to determine whether the technology will, in fact, produce the expected results before the full - scale application of the technology is applied. 2.08 During the final stage of the Bati Raman EOR study, GOT requested assistance not only to undertake a field demonstration test for Bat. Raman, but also lor the development of additional reserves in the Raman oil field, and in the assessment of the production potential of the Hamitabat gas field in Thrace, northwestern Turkey. 2.09 In 1980, a loan agreement was signed between TPAO and the IBRD in an amount of USD 62 million for the implementation of 3 projezts in the Bati Raman and Raman oil fields, and the Hamitabad gas field, as described in the following.  - 21 - Project Description 2.10 The main objectives of the project were to increase Turkey's oil production in a short term by developing the newly discovered oil reserves and to expand Turkey's recoverable reserves and enhance its medium-term oil production capacity. These objectives would be achieved by the implementation of the following 4 subprojects. 1. The Bati Raman Field Carbon Dioxide Injection Project: A field demonstration project of the chosen carbon dioxide injection technology would be conducted on a linmited area of the Bati Raman field for the purpose of evaluating the suitability of this technology for full-scale application, and determining the data required for designing the whole field project. The oroject covered the construction of a 75 km pipeline to transport 50 MMSCFD of carbon dioxide gas from the Dodan gas field to the Bati Raman field; the installation of associated surface systems at both fields; the preparation of 25 wells and the drilling of 5 additional wells in the Bati Raman field; the application of the above mentioned technology and the evaluation of the results. 2^ The Raman Field Development Project: A production project covering the cost of 18 new production wells and associated surface equipment in the northern extension of the Raman field and a reservoir analysis study determining the optimum approach to a secondary recovery program for the field. 3. The Hamitabad Gas Field Stimulation Project: An evaluation of the production potential of the Hamitabad gas field to provide reliable estimates of gas reserves and productivity, and a study to determine the optimum use of the gas.  - 22 - 4. The Management Study: Providing TPAO with assistance for improving its operational productivity and its management organization. III. PROJECT IMPLEMENTATION 3.1. THE BATI RAMAN FIELD CARBON DIOXIDE INJECTION PROJECT Project Management 3.1.1 The project management was entrusted to the TPAO Production Department. To assist TPAO during the material procurement, facility construction and the execution of the project, an engineering agreement was signed between Williams Brothers Engineering Comnpany (WBEC) and TPAO in 1981. Also, a contract was sianed with TULAA5, a local engineering company, to supplement WBEC's work domestically in the respective areas of the field surface facilities design and the main pipeline. Two other contracts were signed with The State Highway Department (TCK) and The State Electrical Institution (TEK) for the construction of 12 kilometers road from the state highway to Dodan plant, ard the supply of 10 MVA electric power at 6.3 KV to the Dodan plant, respectively. During the phases of material procurement ard facility construction, contracts were signed with local and foreign companies to carry out such works. kfter the completion of construction works at Dodan and Bati Raman areas, and before start of the carbon dioxide injection, a new agreement was signed with WBEC in 1985 to provide TPAO with procedures and reccmmnendations for operating the iroject and for recording data and monitoring the project performance. Project Cost 3.1.2 The first Loan Agreement dated November 30,1978 between GOT and IBRD has granted to TPAO in various currenries equivalent to 2.5 million USD to assist in financing the Bati Raman EOR Project. Upon successful ccmpletion of feasibility study and the identification of promising EOR schemes outlined in  - 23 - the final report, IBRD decided to encourage TPAO by a second loan. The Loan 1917 TU dated November 24,1980, totaled 62 million USD which also included refunding the first loan out of the proceeds of the second loan. The original appraisal in the Loani Agreement has been amended twice to reallocate necessary funds due to special characteristics of the project. Annex 1 sets forth the categories of items financed out of the proceeds of Loan 1917 TU. 3.1.3 As mentioned in Article 2.10, Loan 1917 TU also includes the Raman and Hamitabat fields development projects which are shown in categories 1-B and 1-C, respectively, in Annex 1. When the sum of final appraisals for both projects is deducted from the total allocated amount, the actual final appraisal of the Bati Raman Project is 50.2 million USD. Project Disbursement 3.1.4 The amount of money disbursed during the execution of the project by the closing date of December 31,1987 is shown in Annex 2. From the beginning to the closing date, the total of 54 839 472 USD has been spent from the proceeds of Loan 1917 TU. The realization is 89 percent which can be credited as the result of close coordination between IBRD and TPAO. Also, Turkish vendors, particularly civil contractors, played important roles in the execution phase of the project. However, in any case, it can be concluded that the project tasks have been achieved within the budget which is a good indication of efficient work of all the parties involved in the project. 3.1.5 As mentioned above, the categories 1-B and 1-C in Annex 2 also reflect the amounts for Raman and Hamitabat fields development projects, respectively. When the sum of disamounts for both projects is deducted from the total disamount, the balance is the disbursement of the Bati Raman project which was45 529 849 USD by the closing date. Similarly, the remaining amount for the Bati Rdman project was4 670 152 USD as of December 31,1987.  - 24 - Procurement and Construction 3.1.6 The material procurement of the project was made under the IBRD's procurement procedures. The International Competitive Bidding (IC6) was applied to the materials procured. The procurement of specialized natures was also done under limited international biddings. In the initial stages of project execution, some difficulties related with the IBRD's procedures were en-ountered. This was generally due to the TPAO's unfamiliarity to the rules of ICP and was eliminated during the project period. 3.1.7 The project involved the procurement of different types of equipment and materials necessary for the implementation. These are divided into following groups: Dodan-Bati Raman linepipes, gas compression system, Selexol plant, dehydration unit, separators, manifolds, steam generation system, instrument air system, off-gEs disposal system, motor control center, central alarm system, chemical injection system, pipeline appurtanences, field gathering lines, surface and subsurface well equipments, electrical equipment end controllers, water treatment system, and stock tanks. Detailed information about these systems are given in the following section. The material procurement list is also given in Annex 3. 3.1.8 The construction of Dodan-Bati Raman pipeline was completed between May 1982 and July 1983. Following the material procurement, the construction of surface facilities at both fields was commenced in May 1984. This work was completed in October 1985. The construction of main pipeline and surface facilities was accomplished by local contractors. During the construction of surface facilities, some problems related with the scheduling and cost was encountered with the contractor. This brought both parties the court to settle the dispute between TPAO and the contractor. In the period from October 1985 until March 1986 when the gas injection started, maintenance and testing of surface facilities was completed.  - 25 - Surface Facilities Dodan Plant 3.1.9 The Dodan gas-processing and compression plant is centrally located in the field and was designed for the Dodan gas to be sweetened, dehydrated, compressed, and then transported in dense phase to Bati Raman field by pipeline. The plant is capalle of producing the equivalent of 60 MMSCFD of carbon dioxide-rich reservoir fluid from 12 wells,of removing hydrogen sulfide and water, and delivering about 55 MMSCFD or dry hydrogen sulfide-free gas to the Bati Raman pipeline. 3.1.10 Fiberglass reinforced thermosetting resin pipes (RTRP) were used as gathering lines to transport the reservoir fluid from wellheads to the Dodan process plant. The design includes provisions for the addition of heat at the wellhead to prevent formation of hydrates in the flowlines from those wells producing near the hydrate temperature. The fiber-glass pipe is rated for 2000 psi, but the maximum allowable gathering system pressure is 1200 psi as limited by stainless steel 316 L flanges used at the manifold and wellhead connections. All flowlines at Dodan area are 4.5 in. in dianeter, insulated, and buried. They had to withstand extzemely corrosive fluid at 1200 psi, and be installed in rugged terrain inaccessible in some areas to tracked vehicles. These requirements called for a system which did not require cumbersome construction equipment and which could be repaired or expanded using equipment readily available in Turkey. The major concern regarding the use of RTRP was its lack of previous history in high pressure carbon dioxide/hydrogen sulfide/water piping systems; however, it has proven to be corrosion resistant in a wide range of severe applications including downhole service. Another concern was the relative susceptibility of the pipe to mechanical damage, considering the toxic nature of the produced gas. To minimize the potential for damage, the flowlines were buried and stainless steel connections to the wellheads and manifolds were tied in to the fiberglass below grade.  - 26 - 3.1.11 TIle produced reservoir fluid includes free water, carbon dioxide, and other hydrocarbon gases with some other impurities. Free water is removed by means of two-phase separators. Liquid carbon dioxide in the incoming stream is evaporated using waste heat of compression from the pipeline compressors. The gas is delivered to the Selexol plant for the remova'l of hydrogen sulfide. 3.1.11 The produced reservoir fluid includes free water, carbon dioxide, and other hydrocarbon gases with some other impurities. Free water is removed by means of two-phase separators. Liquid carbon dioxide in the incoming stream is evaporated using waste heat of compression from the pipeline compressors. The gas is delivered to the Selexol plant for the removal of hydrogen sulfide. 3.1.12 The Selexol plant consists of two parallel 50 percent trains, each capable of operating independently of the other. This two train configuration also allows an effective gas processing at rates below 60 MMSCFD. Operation at reduced rates is also allowed when either one of the units is down for maintenance. When operating at design conditions the Selexol plant removes about 5 MMSCFD (8 percent) of the carbon dioxide with the hydrogen sulfide. rhis hydrogen sulfide-rich gas is disposed of by burning in an incinerator and venting the sulfur to atmosphere in the form of sulfur dioxide. The pipeline gas from the Selexol unit contains about 25 ppm hydrogen sulfide by volume. After a preliminary evaluation of possible separation systems, th- Selexol system has been selected as the best one for Dodan service with the 700 psi minimum plant operating pressure being a primary consideration. Being a physical solvent, Selexol was able to use this relatively high pressure to achieve a high acid gas loading and to largely regenerate the solvent without heat by flashing the solvent at lower pressures.The chemical solvent performance is independ nt of pressure and is unable to capitalize on the high absorber operation pressure. Another factor leading to the selection of Selexol is its selectivity of hydrogen sulfide over carbon dioxide of a factor of 9-10, capared to 4 for the  - 27 - chemical solvent being considered. Equipment used in the Selexol system did not require extensive use of special materials. The Selexol solvent is both non-corrosive and has dehydration properties. Thus, carbon steel surfaces wetted by Selexol are normally protected by the solvent, and the carbon dioxide gas in the equilibrium with the Selexol has a water dew point below the operation temperature. Therefore, use of stainless steel was generally limited to vapor sections isolated from the main flow, such as bridles and PSV connections, where film temperatures fall below water dew point during cold weather. The one major exception to this was the Selexol stripper and some related equipment. This equipment, which is continuously exposed to saturated steam, carbon dioxide, and hydrogen sulfide, was fabricated entirely of stainless steel. 3.1.13 After sweetening, the gas coming from the Selexol plant is dehydrated in a triethylene glycol system to 4-12 Lbs water/MMSCF to prevent water condensation in the pipeline at operating temperatures as low as 0 degrees centigrades. This dew point specification allowed the use of unlined carbon steel in pipeline construction, provided the line is buried below the frost line. With the exception of the dehydration system, plant facilities consist of two parallel process trains each rated for 50 % of plant capacity. The single dehydration system is rated for fuli capacity, but has spare rotating equipment and high turndown trays in the contractor. This design ensures effective operation at rates as low as 10 MMSCFD and also allows continued, reduced-rate operation in event of failure of critical equipment. 3.1.14 The Dodan gas is compressed in two single-stage, motor-driven reciprocating compressors to 1745 psi. The gas is cooled, metered, and delivered to pipeline at about 1705 psi and 115 degrees fahrenheit. 3.1.15 The production and processing facilities at Dodan are designed to operate at about 700 psi. As pressure falls significantly below 700 psi in the future due to reservoir depletion, it will be necessary to install a wet gas compreSsion  - 28 - system at the inlet to the plant in order to maintain full rate delivery to the pipeline. Such a system would compress the gas from gathering system pressure to 700 psi. Thus, the gas processing and pipeline compression systems would not require modification. 3.1.16 A two-boiler, closed-loop steam system supplies heat for the Selexol plant, glycol system, liquid carbon dioxide evaporation coils, and fuel gas heater. The steam system operates at 250 psi and is capable of delivering up to 30 million BTU/hr. The normal steam demand is about 20 million BTU/hr. 3.1.17 An instrument air system supplies air for instrumentation and controls as well as for pneumatic tools. It includes two 100 % compressors to ensure availability. 3.1.18 The plant, the large portion of which was supplied by Turkish -idustry, was manufactured to ASME codes. It was preassembled and skidmounted complete with local control panels by the suppliers. Normally, the skid assemblies were functionally tested before shipping to the field. Produced- fluid handling facilities exposed to wet carbon dioxide were manufactured of different materials depending on the service. In the plant area, corrosive service equipment and piping were generally fabricated from stainless steel or stainless- clad carbon steel. Thin film epoxy lining and fiberglass pipe were also used in the plant in some low pressure and marginally corrosive services. 3.1.19 Water supply for the plant is delivered fram a pump station located on the nearby Basur river at up to 500 BPD. The water ia stored at the plant and processed in a water treatment system for use as boiler make-up and Selexol process water. The incinerator and steam generators are operated with fuel oil. The daily demand of fuel oil is about 15 tonnes. Primary power of the plant is supplied by the national network. The normal requirement of approximately 5.8 MWA is delivered to theDodan plant site at 6.6 via a transmission line and  - 29 - substation. Emergency back-up power is supplied by an on-site 125 kVA diesel-driven generator. The Dodan-Bati Raman Pipeline 3.1.20 The 10 in. diameter, 80.5 km Dodan-Bati Raman pipeline was designed in accordance with ANSI B31.8. It is constructed of API-5LX X60 carbon steel pipe with wall thickness of 0.297, 0.307, and 0.438 in. The normal in.,et and discharge pressures are 1700 and 1550 psig, respectively. The maximum pressure is 2500 psi and occurs at the pipeline low point during cold weather operation. The pipeline route passes in the vicinity of several smaller heavy oil fields operated by TPAO. There are five takeoff points so that the pipeline can supply these fields in the future. Other features of the pipeline, which varies in elevation from 1700 to 3700 feet above sea level, are three river crossing, three pipeline valves, blowdown stations, and pigging and metering facilities. Bati Raman Field 3.1.21 Much of the existing equipment at Bats Raman field was incorporated into the design to minimize capital cost while providing operational flexibility consistent with the needs of the project. The western test area of Bati Raman was designed for cyclic carbon dioxide injection and production of 33 demonstration wells using transfer pump stations 3TP1 and 3TP2 as the centers for operations. Existing facilities at the field consist of numerous remote manifolds which feed into trunklines to satellite pump stations. 3.1.22 The western area carbon dioxide injection system supplies approximately 20-25 MMSCFD of gas at about 1500 psi wellhead pressure to the demonstration we'lls operating in the injection mode (up to 17 wells simultaneously). The production facilities handle the oil, gas, and waiter from up to 17 demonstration wells in the producing mode. The 3TP1 and 3TP2 pump staticns include facilities to evaluate  - 30 - individual well performance, to remove produced water and gas, and deliver the oilto theAPI main pump station fcr additional dehydration and desalting. The gas produced from the demonstration area is vented to the atmosphere. 3.1.23 Ideally, the gathering lines at the Bati Raman field should be internally lined and use stainless steel valves and fittings to withstand corrosion. They should also be insulated to prevent heat loss and; consequently, high oil viscosity. Without the use of lining and stainless steel fittings, the flowline might be susceptible to rapid corrosion. Without pipe insulation, high oil viscosity during winter operations might limit flowline capacity, and therefore, well production poential due to high flowline back pressures. Owing to the expense of installing a new gathering system, only a few of the new lines were included in the proposed design. The performance of these lines is being monitored to determine if a completely new gathering system is justified. In the production system, where operating pressures are normally be less than 100 psi, there is negligible hydrogen sulfide, and initially, water is not present in the gathering system as a continuous phase. Corrosion-resistant materials specified for Bati Raman surface facilities consist primarily of fiberglass flowlines (6 and 8 in. at 700 psi) on selected wells. Also, thin film epoxy phenolic coating was specified for all new in-plant piping and separators. This includes all systems handling water-saturated carbon dioxide gas. In the Bati Raman area, most of the existing carbon steel stock tanks, pumps. flowines, and oipelines which were predominantly in liquid service are being protected by injection of liquid corrosion inhibitors. Project Planning 3.1.24 In spite of uncertainties involved, the results of initial comprehensive studies showed that a substantial quantity of oil could be recovered by cyclic Dodan gas injection possibily followed by water flood. The best way this could be tested was in the field.  - 31 - 3.1.25 It was decided that a pilot be initiated to test oil recovery by a cyclic Dodan gds injection, and depending on early performance of the reservoir corrective measures would be taken. The general area chosen for this demonstration project is in the western part of the field, encompassing about 1200 acres with 33 adjacent wells drilled in a five spot pattern (Figure 1). To begin the operation, half of the wells (first, third, and fifth lines of wells shown in Figure 1) were planned to be used as injectors, and the other half as producers. After about 3 months, when bottomhole pressures reach about 1800 - 2000 psi, injectors were planned to be converted to producers, and producers to injectors. This mode of operation, i.e. out of phase stimulation of adjacent wells was based on the assumption of a single porosity system. When a well experiences the injected Dodan gas, neighboring wells were planned to be put on production so that the injected gas could be readily move out from the well and finger through the oil allowina diffusion. In this case, gas break-through should not occur for a long time. If the reservoir in all or in part of the test area is closer to a dual-porosity system, then break-through in the producers should occur early during the scheduled first injection cyc.le. In that case, the operation was planned to be converted to in-phase cyclic stimulation of one half of the area at a time. Thus, while the eastern half was being stimulated, the western part should be on production, and vice versa. Durirg the implementation of the project, both conditions, buc mostly the existance of the dual porosity system, have been experienced in the pilot and vicinity areas. This will be discussed in the following sections. 3.1.26 Initial work for the start uD of the Bati Raman EOR demonstration project was commenced in November 1985, and consisted primarily of establishing project monito-ing procedures for the start up, organization and implementetion of the computer database system and the reEervoir simula:or. Past performance and core and log data of the Bati Rainan wells were reviewed in 1985 for the development of a reservoir  - 32 - model suitable for the simulation work. Initial simulation work consisted of single-well huff and puff simulations. This early work demonstrated the importance of buildirg reservoir pressure to a level where carbon dioxide solubility reaches significant levels. It also showed that ap?lication of the huff and puff process would require contai:.ment of the carbon dioxide within the demonstration area. 3.1.27 After the new agreement was signed with WBEC for monitoring the pilot application, J. S. Nolen and Associates's Dual VIP simulator was used for simulation studies. This simulator is a tnree-dimensional, three-phase, dual porosity, "black oil", reservoir simulator which accounts for diffusion of gas between fracture and matrix and the 3olubility of carbon dioxide in water. Simulation runs were made on a Microvax II computer for about a year. and in December 1986, the simulator was installed on the VAX 11/750 computer located at the TPAO Production Group office in Ankara. 3.1.28 The simulator used in the simulation studies made in 1979 was not a dual porosity simulator. In that study, reservoir fracturing was accounted for in a simplistic and non-rigorous manner. It was not until several years after the first simulation study was performed thLf technology for rigorous, dual porosity simulation became available. Also, the simulator used in the 1979 study was not fully implicit and therefore could not run radial, single well simulations in an efficient manner. As a result, all studies of single well processes such as huff and puff, were done in cartesien coordinates. Results obtained when simulating a single well in this way differ significantly from results obtained using current simulators. 3.1.29 The reservoir simulation technology has been improved considerably since the previous study was made. An entirely new generation of reservoir simulators was developed during this time. Also, considerable advances in methods and techniques for applying simulator results was made. Therefore, it should not be surprising that the results and conclusions obtained in the final simulation studies are different from those of  - 33 - the previous ones. Well Design 3.1.30 rhe principal design criteria for the wells in Dodan area was the extensive corrosive environrent. The production casing is 6 5/8" J-55 and was cemented is 2 7/8" in grade L-80 with plastic coated internals. The producing formation was isolated by hydraulically set p3rmanent packers at the top. Wellhead, gauges, needle valves, adjustable chokes and master valves had been trimmed for hydrogen sulfide and wet carbon dioxide services. 3.1.31 In the Bati Raman pilot area, 29 of the pilot wells were completed with single tubing and packer. Both injection and production is being done through the 2 7/8" tubing. Following the injection cycle, the well is allowed to flow through the tubing. When pressure is depleted and the wells stop flowing, the pump and rods are run into the well. At the end of the production cycle, the pumps and rods are pulled out and the injection tree is re-installed. For this switching operations a pulling unit is required. In the pilot area, a dual tubing completion, one for injection and the other for production, was designed for the remaining 4 wells. For existing 6 5/8" casing wells, 2 7/8" and 1.9" strings were run with dual hydraulically set packers. By this design, the rods and pumps do not needed to be used for fluid entry surveys. Also, in the pilot area, 5 wells in key locations are equipped with pressure - temperature sentry systems. All pilot wells have 2" bottomhole gaseous puirips with D-320 pumping units.  - 34 - 3.1.32 Corrosion is not significant during the injection cycle as the Dodan gas should be dry and virtually free of hydrogen sulfide and other contaminants. Although the downhole equipment is subject to a wet carbondioxide environmen;. during production, the tubing is not plastic coated since the coating is eroded by rod action during the pumping cycle. However, during the project execution, corrosive acti.ons shown in downhole equipment has not been found serious. An oil soluble, water dispersible corrosion inhibitor is being squeezed after each well is put on production and just before the start of the first carbon dioxide injection cycle. Each producer is then resqueezed before pumping operations start. In the pilot area, all wellheads, gauges, valves, needle valves, tubing hangers and master valves were trimmed for wet carbon dioxide service. Project Execution 3.l.33 The Dodan - Bati Raman pipeline was first used to transport the Dodan gas for the pilot project application in the Bati Raman field on March 19,1986 after the surface facilities at Dodan plant had been tested for maintenance and performance. However, particularly due to the malfunctioning of the main line compressor, continuous gas injection could not be achieved and the injection was stopped 6 weeks later. During this 6 - week period, a total of 520 MMSCF of gas with 12 MMSCFD average flow rate was injected into the Bati Raman reservoir through 16 injectors located on the first, third and fifth rows  - 35 - in the pilot area (Figure 1). Injection was restarted in July 1986 following the completion of repairs and restoration on compressors and other facilities at the Dodan plant. But, because of necessary modifications of the compressors and the incinerator, the injection was had to be stopped again several times , totally about 6 weeks, until November 1986. Between August and November 1986, a continuous 2.5 - month injection was firstly accomplished with an average rate of 22 MMSCFD. In a 3 - week stop in November 1986, the compressors and other systems were readjusted and the whole system at Dodan plant was reexamined. Starting from this date up to now, continuous injection was accomplished at an average rate about 20 MMSCFD, except for two stops in October 1987 and January 1988 for maintenance and some repairs. Also, for testing purposes, the injection rate was increased to 48 - 50 MMSCFD in February 1987. The injection profile of the project is shown in Figure 3. In September 1988, the total amount of gas injected into the Bati Raman reservoirs was 14 MMMSCF since the beginning of injection. Currently, injection continues at a rate of 18-20 MMSCFD. 3.1.34 On the other hand, as far as the pressures and flow rates of Dodan wells are concerned, all the observations were positive during injection of gas at an average rate of 3 MMSCFD per well and a maximum rate of 6.5 MMSCFD per well into 10 Dodan wells that have an average of 1000 - 1100 psi dynamic wellhead pressures.  - 36 - 3.1.35 Before start of the injection, wellhead pressures of 18 injection wells were in the range of 300 to 500 psi in the pilot area. After the injection was started, it was observed that the injectivity of wells located in the eastern part of the pilot area was comperatively higher (1.1 MMSCFD) than that of the wells located in the western part (0.4 MMCSFD). After observing gas coming out of the wells * 171 and v 197 in the pilot area and the vicinity well a 112 at the east of the pilot area, the injection into 9 wells located at che east half of the pilot area was temporarily stopped in order to prevent the migration of gas to where the reservoir pressures was relatively low. The Kh factor was improved in the central area of the field. Meanwhile, since the designed injection pressure of 1500 psi could not be maintained earlier, the injectivity of wells located in the west part of the pilot area was improved to a reasonable level as those wells in the eastern part were closed to injection. After obtaining a sufficient pressure level,the wells previously closed were put on injection again in October 1986. Since then, the injection has continued mos2.-y into 18 wells in the pilot area, and in that period 9 injectors were put oir production mode according to the designed "huff and puff" application. 3.1.36 In September 1986, in order to prevent the migration of injected gas to the central area of the field and in order to have a more effective injection in the pilot and vicinity areas, water injection was started with the rate of 5000 BPD into the wells w72, w33, v99, and f69 located on the boundary line between western and central areas of the field (Figure 1). By this applicatior, the injected water formed a barrier between the central and western areas preventing injected gas from escaping fran the western to the central area of the field. The water injection continued for 9 months until May 1987,and in that period an effective pressure barrier was obtained in the mentioned area which was about 900-1200 psi at the west of injectors #72 dnd r33, and 800-1200 psi at the south of injectors 169 and p99. No gas break through was observed and also the injected water had partially effected the central area of the field resulting in production increases in some wells. During that period, a total of 950 000 Bbls of water was injected into the Bati Raman reservoir.  - 37 - 3.1.37 During the project application, pressure develcpment has been observed by periodic surveys in the pilot area and surrounding wells. Wells #107 and #118 located on the injection lines and wells 172, 182 and 194 located on the production lines were initially selected as key wells for continuous pressure and temperature surveys. Pressure profiles recorded by these wells is shown in Figure 3 except well #107 at which the pressure/temperature sentry system failed. The bottomhole pressures that were less than 400 psi in early stages of the injection scheme in wells f172, 0182, and #194 rapidly increased after the continuous gas injection has effected their areas starting from October-November,1986. The pressure in wells 172 and 194 has reached levels up to 1350 psi, and in well #182 the pressure increased to 1850 psi in March 1987. As can be seen in Figure 3, this was a result of high injection rates tried only for testing purposes for a 3-week period. The primary purpose was to test the injection capacity, but it also showed that higher injection rates didn't result in significant changes in production rates, so the injection rate was maintained at about 25 MMSCFD between March and July 1987. Due to the increase of the back produced carbon dioxide gas, the injection rate wa3 decreased to 15-18 MMSCFD between July 1987 and March V 1988. After that date a good injection/production balance has been achieved by an injection rate around 20 MMSCFD. As also seen in Figure 3, the bottomhole pressure in well #118 which was an injector at that time increased rapidly to 2500 psi during the first few months of the injection. Then, it decreased to 1000 psi due to interruptions of injection until August 1986 when a continuousm injection was achieved. The pressure increased again starting from that date, and in October 1986 it reached 2300 psi. This well was put on production in November 1986, and therefore the pressure profile shows a decline after a soaking period from that date until February 1987. Since then, the well has been on production and has about 600 psi dynamic bottomhole pressure. The static bottomhole pressure of key wells was calculated to be about 1700-1800 psi when the reservoir was pressured up by continuous injection. This pressure level was achieved in March-June 1987 and it has been maintained since then. The gas that was injected in almost all directions in the central and more heavily in the eastern  - 38 - regions of the pilot area generally tends to migrate to the structually higher portions of the reservoir where large fractures axit, causing high pressures in those portions. This situation, in one respetct, can be verified since the recoveries of the production wells in the pilot area increase with the increasing rate of gas backproduction. For example, the production in well #173 located in the center of the second line of the pilot area has exceeded 300 BPD with a gas backproduction of 1 MMSCFD; and over 200 BPD of oil has been produced with a gas backproduction of 1 M.MSCFD from well #193 located in the fourth line of the pilot area. Starting from March 1987, a high pressure area was observed initially at the second and fourth line producers of the pilot area as recorded by the key wells. Also, static bottomnhole pressure measurements made in surrounding wells showed high pressure areas initially at the first line of the southwest pilot area. On the other hand, pressures in the areas where wells =112 and =68 are located east of the pilot area and wells =153 to =102 are located as a line north of the pilot area remained low. This was explained possibly by ar. existence of a low permeability barrier between these areas. 3.1.38 During later stages of injection, it was observed that high pressure areas extended to the all surrounding parts of the pilot area. In the beginning, it was expected that the production increase would take place only within the drainage area of the injectors by the application of "huff and puff" method. However, as the pressure i..vreasa was continuing and spreading over the whole pilot area and most of the surrounding wells, it was seen that the "huff and puff"a'zolication turned into a more effective gas drive. This is one of the most important results obtained by the field application cf the Dodan gas injection. 3.1.39 Most of the production wells that were previously shutdown because of the low reservoir pressure in the pilot and vicinity areas were then put back to production as fluid levels increase due to the gas injection. Hence, the  - 39 - total production rate of the 3 main gathering systens (3TPI, 3TP2, and AP1) of the oilot area and surrounding wells increased from under 300 BPD before March 1986 to over 1000 BPD in April 1986 because of the additional producing w.ells being put on production with higher fluid levels. Oil production and gas backproduction rates and gas-oil ratios of these 3 gathering system are shown in Figure 4. In Figure 5, oil production and gas injection/backproduction of the "test field" (Pilot and the vicinity areas) are also shown in daily and cumulative figures covering all 3 gathering systems. As seen in Figure 5, since the injection was temoprarily stopped for 3 to 4 months until September 1986, the production of the test field declined to the 800-900 BPD level;once the injection was restarted the production Luoe back to 1200 BPD in October 1986, 1400 BPD in January 1987 and over 2000 BPD in March 1987. Production increasescontinued and reached, 3000 BPD in July 1987 (Figure 5). Since then, over 3000 BPD production was obtained from the test field and in September 1988 it reached 3700 BPD. 3.1.40 The number of production wells from which the injected gas is back produced has increased along with the injection. Although there were only 2 to 3 wells that back produced gas in April 1986, the number of gas back producing wells have increased to 12 in the pilot area and 5 in the vicinity area in March 1987. Ir. September 1988, the gas back production spreaded over all producing wells in the pilot area and 15 vicinity wells. The first gas break through was observed in well #112 and then well p120 which are located where Kh contours have higher values. Although a gas flow in the direction of those wells was observed in the beginning, there was not a considerable gas flow from the other vicinity wells later. Starting from October 1986, the gas back production rate reached to 1-2 MMSCFD in December 1986 and exceeded 4 MMSCFD in March 1987. As the oil production increased in the pilot and vicinity areas as a zesult of the continuous injection, the gas back production also increased (Figures 4 and 5). It was over 10 MMSCFD in September 1987. Considering the maximum capacity of the separatinq system in 3TP1 and 3TP2 stations which is 12 MMSCFD, some adjustments were made in the injection programme. After the injection rate was fixed at around 18-20 MMSCFD, the gas back production rate was  - 40 - maintained at less than 10 MMSCFD since last September. It was 9.4 MMSCFD in September 1988. After the installation of re-cycling of the project extension, re-injection of back produced gas will be accomplished. 3.1.41 The aim of "huff and puff" application for the pilot prolect was to increase the bottomhole pressures of the wells up to 1300-2000 psi level and put the wells into production after a soaking period. Once the pressure had reached the desired level, the injection wells 150, 118, 214, 156, 117, 116, 115, 103 and 108 were put into production between November 1986 and February 1988. General production performance of these wells is as follows. As the wells are put on production, the gas starts to be produced at a rate at about 1 MMSCFD, and within the following few days oil production starts together with the gas in a flowing phase. The well generally flows about 1 to 2 weeks depending on the pressure decrease. The average recovery of the wells is about 100 BPD in this period. After the well is placed on pump, the production continues with the rate of 50-60 BPD for 2 to 3 weeks, then it gradually decreases to 25-30 BPD level within 2-3 months. O.n the other hand, the GOR is very high in the beginning when the well flows. In the pumping mode, it decreases below 10000 SCF/STB, and continues to decrease gradually until i: reaches almost a constant value at about 3000 SCF/STB in few months. The only exception to this behavior is well 118. When this well was first put on production, a high flow rate of gas with no oil was observed for about 2 weeks. Then the well started to produce oil in increasing rates up to 60 BPD. In September 1988, this well has almost 2 years production life after the injection mode and produced at a constant rate of about 50-60 BPD. Although the well wasn't put on injection mode again, a continuous oil recovery at about the same rate was obtained. This behavior demonstrates the areal pressurized reservoir effect rather than the "huff and puff" effect. 3.1.42 The total production of Bati Raman field has noticeably increased when compared to the primary decline of the field for mid 1982 due to the water injection applied in various  - 41 - patterns in between 1972 and 1981; however the production returned to its original primary decline after 1982 (Figure 6 and 7). The daily production of the field which was 2500 BPD in 1982 had declined to 1500 BPD lust before start of gas injection. The total field production has also risen along with the gas injection becoming 1750 BPD in May 1986, 2000 BPD in September 1986, 2500 BPD in February 1987, 3000 BPD in March 1987, 3500 BPD in June 1987 and 4000 BPD in July 1987. Since then, the production capacity of the project area has been at the same level, and the total field production has averaged around 4000 BPD. In September 1988,01o1 production reached 4600 BPD. If the gas injection had not been implemented, the production decline curves for the field ssiow that the primary production would have been 1000 BPD as of September 1988. In consideration of that fact, there has been. a production increase of 3500 BPD as of September 1988 which corresponds to a 4.5 fold increase in production. 3.1.43 In Figure 8, the Bati Raman project injection-production efficiency is reptesented. In this figure, the incremental oil produced from 3 main gathering systems of the project area is plotted vs. the total gas injected since the beginning of the ga3 injection. It can be deduced from the slope of this curve, especially upon considering the production increase since the second half of 1987, that 5000 SCF of gas is needed to be injected for 1 STB of oil produced. This ratio is two times smaller than the average figure of 1 STB/10000 SCF, given by the industry. 3.1.44 Since the beginning of the project application, all field data has been recorded by the Ingres Data Base System to the VAX 11/750 Computer System installed at TPAO Production Group in Ankara. This evidently provided easiness to the project monitoring. Figures 9 and 10 are given as examples showing injection and production histories of 2 wells, one injector and one producer, respectively. Similar figures for all of the wells existing the pilot and vicinity  - 42 - areas can also be^n obtained. Figure 11 has been used to follow up the final status of wells existing in the 3 main aathering systems. Also, distributicns of pLoduced oil, pressure and GOR contours in the project area can be obtained in X,Y,Z coordinates systems. Figure 12 thru 14 represent these distributions. 3.1.45 The Bati Raman reservoir simulation model studies showed that more economical recovery could be obtained by a pattern drive process utilizing both carbon dioxide and water. it was proposed by our consultant (WBEC) that the benefits of a WAG (water alternate gas) drive process were dependent upon the water imbibition characteristics of the Bati Raman reservoir rock. As the available laboratory and field data we,e insufficient to determine these imbibition characteristics, it was decided that the best way to evaluate these characteristics was with a field pilot.For this purpose,2 gas injection wells (#107 and *148) at the west side of the pilot area were chosen as water injectors (Figure 1). In March 1988, water injection st2rted in these 2 wells at a total rate of 500-600 BPD. Wells s176 and #177 east of the injectors were chosen as observation wells. In April-May 1988, the effects of water injection were started to be observed in these 2 wells with increasing water cuts. Although the total injecticn rate was dropped to 300 BPD in June 1988, the increase in water cut continued up to about 70-80 %. In September 1988,the total amount of injected water was 130000 Bbls, however there was no remarkable increase in the oil production of the observation wells.Before the start of the water injection, production rates fram the observation wells had reached over 100 BPD due to the injected gas which was about 1 MMMSCF in total from wells #107 and p148. During water injection, high water cuts caused a drop in production of the observation wells down to the 40-50 BPD level. On the other hand, in well t175 which is located north - east of the water injector #148, a production increase above 100 BPD was observed after the start of water injection. This increase could be due to the effects of gas injectors #214 and #150 located at north of this well, however  - 43 - it. is also believed that a sufficient pressure increase might be supplied to the vicinity area of this well by the effects of w;ater injection from well 0148. This is not proved yet, because no water has been produced from 0175. The pilot application of the WAG project has been continued, however field results obtained up to now suggest that it has not yet shown promise. Performance Analysis 3.1.46 Conventicnal use of carbon dioxide for improving cil recovery has been mostly confined to miscible applications where the displacement of crude oil from pore space in rock is achieved by a solvent action that prevents formation of interfaces between driving and driven fluids. However, in the Bat. Raman field, such miscibility could not be obtained and has not been a factor in increased oil recovrery. Rather, the enhancement of oil recovery should result from the high solubility of carbon dioxide in iow-gravity oil. The initial model studies had shown that the major task of the project was saturating substantial quantities of oil by cyclic injection of Dodan gas. The process has began by repress3ring portions of the reservoir withl DDdan gas allowing it to diffuse into the oil it contacts. Oil recovery by Dodan gas diffusion is the primary mechanism in which Dodan gas can dissolve in oil, causing oil swelling and viscosity reduction. But, diffusion requires a long time iefore a substantial amount of reservoir oil can be treated. Before diffusion can work as a practical means of treating large volumes of oil, it is necessary that the injected gas effectively penetrate the reservoir and provide a large areal contact. 3.1.47 Simulation of past performance data of the. field and water-flood pilots had shown enough evidence that a dual- porosity system dominates at least in certain parts of the reservoir while a combination of duel and single porosity systems can be effective in other portions. The nature of the dual porositysystem with its highly conductive and low fracture network, is conductive to an effective penetration  - 44 - of the reservoir as Dodan gas moves through the fractures. Thus, a dual porosity system has the basic requirement for success of oil treatment by Docan gas. In a single-porosity system where the fracture network is absent, extensive fingering of Dodan gas through the oil would be necessary to provide the large areal contact needed for diffusion to be effective. 3.1.48 Once the reservoir is pressurized up to 2000 psi and saturation is achieved to a considerable extent, the wells are depleted and a typical solution gas drive performance is expected. In the "huff and puff"model, the pressuring and depressurina cycles, each aiming to contact further virgin oil, should be repeated several times ur.til all recoverable oil is produced Initial model runs had indicated that up to 25 % of initial oil in place could be recovered by cyclic methods depending on unfavorable or favorable selection of parameters involving some uncertainties, well spacing, and number of cycles. This would also result in about a tenfold reduction in oil viscosity and approximate 20 percent increase in volume at the res_rvoir pressure of 2000 psi. After a large volume of oil has been saturated, operating alternatives should include a) Continuing with cyclic Dodan gas stimulation, or b) Water flooding the saturated oil. However, during the field application of the project, well performances have sh3wn that the most of the production increase was coming from wells benefitting from a 'gas drive" mechanism rather than cyclic injection. This was also proved by recent model runs matching with field results. These studies show that a drive pzuc,.bb using only carbon dioxide results in an ultimate recovery of less than 10 % of the original oil in place. This behavior of the Bati Raman reservoir led us to make the necessary changes in the injection scheme considering the gas drive pattern. 3.1.49 The Bati Raman project has been monitored using reservoir simulation models. The objectives of simulation studies were to develop an operational plan optimizing oil production at the Bati Ramr-an field by Dodan gas injection,  - 45 - to use a reservoir simulation model to match reservoir response to gas injection in the pilot area and to predict future performance. J. S. Nolen and Asscciates' Dual *'IP simulator has been used in these studies. The simulator is a three- dimensional, three-phase, dual porosity, "black-oil", reservoir simulator which accounts for diffusion of gas between fracture and matrix and the solubility of carbon dioxide in water. A radial geometry was used for analyzing the "huff and puff" performance, while an 1/8 element of symettry of a 5-spot pattern was used to study the gas flood performance. 3.1.50 General characteristics of the "huff and puff' cycles were that when the well is put on the "puff" cycle, the flow stream is only gas initially, followed by increasing oil and decreasing gas production, as pressure around the wellbore decreases quite quickly. A radial model match of this performance (Figure 15) indicates that the fractures around the wellbore are fully saturated by gas when the well is initially put into produztion. It is only after all of this free gas is produced, and relative permeabilities to oil in the fractures are restored, the oil production starts. Thus, the oil is fed to the fractures from the matrixes. The permeability of oil in the matrix determines the Lil production at the wellbore. Because the pressure declines by this time around the wellbore, also causing dissolved gas to liberate, the oil production from the weLls remains at 20-40 BPD level. 3.1.51 A significant observation at the field was made that a considerable amount of oil was being produced due to the flooding effect of gas. Therefore, the gas-flood performance in a 5-spot pattern was studied. Producing well #182, with 4 injectors #116, 117, 108, and 109 (Figure 1) was simulated. The basic results of this work is as follows. a) The reservoir behaviour is clearly dual-porosity. b) The observed breakthrough on production wells is after-3-6 months of effective injection. This can only  - 46 - be explained by diffusion of carbon dioxide into the oil. If there were no diffusion, the gas would have broken through in less than a month. The model studies also indicate that gas solution in the matrix blocks is a function of the "diffusivity constant" and the "matrix and fracture exchange coefficient" which is a function of matrix block sizes (which is also indicative of fracture intensity). As the diffusivity constant is more or less fixed by laboratory measurements, it becomes a matching parameter. c) In the 5-spot pattern simulation study, the break- through times, bottomhole pressure, gas and oil production rates were matched. This match explains the reservoir recovery mechanism as follows. The injected gas initially comes into contact with the oil in the fractures and dissolves in it. Meanwhile, the gas also diffuses into the matrix blocks. However, since tnis is a slow process, the oil in the matrix blocks always remains unsaturated. As the oil in the fractures becomes saturated, the free gas saturation builds ip in the fractures and then the gas/oil ratio increases. After this 71oment, the only thing that stops the gas coming to the wellbore is its diffusion to the matrix blocks. Therefore, after some tirre, most of the gas is cycled, however the oil production is still considerable (Figure 16). After gas b.-eakthrough, the pressure in the reservoir goes to a declining trend causing the fracture solution gas to decline too (Figure 17). When the injection rates are decreased the gas/oil ratio is improved immediately (Figure 18), but too much lowering will cause the pressure to decline. In this case, there should be an cptimum injection rate. G) The results of simulation model studies show that economic oil can be recovered by optimum injection rates for a long time. Field results obtained later provided evidence that the model is succesfull in predicting the gas-flood performance. Project Expansion 3.1.53 When all "huff and puff" wells a-ere compared with the other production wells in the pilot and vicinity a-eas,  - 47 - it was observed that almost all of the wells in continuous production modes have had much higher recoveries. As of Se.:ptember 1988, the incremental oil recovered by the imple-rontation of the project was over 2 RMSTB. Until that date, only about 10 % of the production increase has been obtained from the "huff and puff" wells and 9G % from the other producing wells. After 1 year of field experience following pro,ect implementatic it was finally interpreted that the production increase was comirg from wells benefiting from a "gas drive" mechanism rather than a cyclic injection. Reservoir model studies as mentioned above also matched with the field results that the prevailing recovery mechanism in the project area was gas drive. 3.1.54 Field results also showed that the maximum oil nroduction of the field by the project implementation at the west part of the field would be around 4000-5000 BPD. To accelerate the oil production, it was ccnsidered that the best way would be the expansion of the project area towards the middle of the field. 11out a 10 fold production increase at the project area encou.aged us to make this step. Considering the gas drive mechanism, necessary changes were made in well patterns and a new area covering about twice as large as the existing project area (Figure 19) was planned for the project expansion. In this application, all injectors were considered in continuous injection mode shown on Figure 19 which required 36 producers to be converted to injectors. Also, in the extended area, new producers were planned to be drilled. Sixteen wells,4 in 1987 and 12 in 1988, are already drilled and completcd. For 1989, the drilling of 14 new wells is planned. In addition, workover operations have been carried out in all existing wells in the expanded area. The large scale commercial expansion of the Bati Raman project covers new injection and production lines and manifold systems; additional separating systems having the capacity of processing the back produced gas from the expanded area, and re-cycling compressors and dehydration systems in order to recompress this gas back into the reservoir. TPAO is now in the construction and procurement phase of above mentioned work. The large scale project is planned to be completed in 1990. By the implementation of new project,  - 48 - oil recovery from the total field is estimated to increase to around 10000 BPD. Conclusions 3.1.55 Implementation of the Bati Raman EOR project has reflected general characteristics of a field pilot application as being either the reservoir behaviour or the problems encountered in surface systems particularly in early stages. The heterogenous structure of the Bati Raman reservoir and especially the thermodynamic conditions of carbon dioxide have formed a rather complex behaviour for the operating systems. The information gained during project implementation has guided us for the future operations of the project. As of September 1988, the general conclusions of the Bati Raman EOR project are as follows. a) The amount of cumulative gas injected into the reservoir is 14 MMMSCF, averaging about 20 MMSCFD for the operating period. Also, approximately 5 MMSCF of gas has been back produced with the oil from the pilot area. This represents about 35 % of the total gas injected. By the application of project expansion, the back pioduced gas is planned to be re-injected into the reservoir by the use of re-cycling compressors. b) Positive results of the pilot project application have being observed starting in March 1987. The total production rate of tne 3 main gathering systems (3TP1, 3TP2, and AP1) which was about 300 BPD before the start of the gas injection has exceeded 3700 BPD. The total production rate of the field has reached over 4600 BPD whereas it was estimated to be around 1000 BPD on primer production This corresponds to a 4.6 fold increase in production. The total amount of incremental oil produced since the start of the project application is over 2 MMSTB. When the incremental oil produced from thc 3 main gathering systems is compared with the total gas injected, it can be concluded that 5000 SCF of gas needs to be injected for 1 STB of oil produced.  - 49 - c) Field results showed that m)st (about 90%) of the observed production increase is coming from wells benefitting from, a gas drive mechanism, rather than the cyclic injection wells. Even the average well performance is clearly better in the former group of wells. It is observed that the injected gas moves to the upper parts of the Bati Raman structure creating a gas flood effect. In general, early breakthrough has not been observed except in a few wells due to the solubility of the injected gas. d) Simulation studies also confirmed that the gas drive yields a higher recovery than "huff and puff". e) Field results of the applied WAG project show that it has not yet shown promise. f) The above results of the Bati Raman EOR pilot application indicated that the expansion of the project could be economically attractive. Considering the gas drive production mechanism, a large scale EOR project (covering twice the area of the existing project area) is planned. g) During the project application, 16 new wells, 4 in 1987 and 12 in 1988, were drilled and completed in the Bati Raman field. Stimulation operations were also applied to all iiijectors and producers in the project area. For 1989, the dri lling of 14 new wells are planned. h) By the successful application of the Bat. Raman project,it can also be concluded that TPAO engineers and technicians working on the project have had opportunities to improve their technical skills and capabilities by attending courses and seminars, and mostly by working with responsibilities in their areas.  - 50 - 3.2 THE RAMAN FIELD DEVELOPMENT PROJECT 3.2.1 Raman oil field is the second largest oil field of TPAO with 600 MMSTB oil-in-place and 60 MMSTB recoverable oil. Raman oil has a grality of 180 API. Heavy oil, the faulted and naturally fractured nature of the reservoir, and the existence of active bottom water complicate the production characteristics of the reservoir. The Raman Field Development Project was initiated to investigate improving the oil production fron this field by conducting an extensive study on the reservoir and production characteristics of the field together with the conduction of an EOR Study to determine the EOR method applicable for the reservoir. Later, the reservoir and EOR study of Garzan field was also included to the scope of the above mentioned project. Garzan-B and Garzan-C reservoirs contain 300 MMSTB of 240 API gravity oil with 42 MMSTB of recoverable reserves. These fields have been subjected to waterflooding since 1960. 3.2.2 In November 1982, an RFP was prepared and forwarded to six prequalified reservoir engineering consultant companies. Five companies forwarded their proposals to TPAO. An evaluation committee worked on the proposals and forwarded the evaluation report to IBRD for concurrence in April 1983. Upon the concurrence of IBRD, the bidder with the highest points,Intercomp, was invited to TPAO for the final contract negotiations. After the completion of the contract negotiations with 7ntercamp for a "Ccoparative EOR Study for Raman and Garzan fields", the final agreement was signed by both parties in September 1983, and the project was officially initiated on Septemberl9,1983 in Calgary. 3.2.3 The project started with data gathering. All related data were reviewed by the project teams of TPAO and Intercomp in TPAO's Production Department, Research Center, and at the Batman field site. Data necessary for the study were copied by Iptercomp and other required data which was not readily available were recommended for preparation. After the data  - 51 - collection, the data editing phase started. In the meantime, necessary laboratory tests were determined and designed together with Intercomp's engineers in TPAO's Research Center. TPAO's Research Center conducted various ccre and PVT tests for the study. Tests for determining the diffusivity of CO, in Raman and Garzan oils were also conducted by the. TPAO Research Center. Only a steam flooding and an insitu combustion test were conducted outside of Turkey by a laboratory in Calgary, Canada (HYCAL LAB). 3.2.4 The project consisted of the following phases and was to be completed in 18 months. A) Geology. B) Reservoir Engineezing. a) History Match. b) Interim Report. c) Secondary Recovery Studies. d) romparative EOR Study. e) Screening EOR Alternatives for Field Applications. f) Prefeasibility Report Presentation. At this stage of the project, the Intercomp company was merged with the Scientific-Software Company and changed its official name to Scientific Software-Intercomp (SSI). The merge did not effect TPAO's project. 3.2.5 Participation of TPAO engineers in the project began with the geological phase and evaluation of the gathered reservoir and production data. A geologist and a reservoir engineer worked with SSI on this phase. Two other reservoir engineers took part in the history matching phase of the study. 3.2.6 The detailed geological studies were finished and the geological and petrophysical reports of both Raman and Garzan fields were forwarded to TPAO for approval in January, 1984. These geological models were the basis of the reservo-- numerical model initilization.  - 52 - 3.2.7 The reservoir engineering phase began with the evaluation of all production, reservoir and well data. PVT and rock data were also reevaluated and finalized forms were prepared for the use in numerical model runs. Upon the investigation of reservoir characteristics, appropriate numerical models were selected to study the reservoirs. A conventional black oil simulator with a feature to study faulted reservoir (BETA II-F) was selected to study Garzan-B and C reservoirs. For the Raman Field, the indication of an effective natural fracture network in the reservoir and areal changes in the oil properties necessitated a more comprehensive model, KAPPA, SSI's black oil model with features to simulate naturally fractured, faulted reservoirs containing oil with spatially variable properties. 3.2.8 In the history matching phase of the Garzan-B and C fields simulation study, some problems were encountered due to the nature of the injectivity data. Injectivity tests were required by SSI, and the results of these tests were utilized in redistributing the total injected water rates to individual wells. Finally the history matching of Garzan fields was completed and studies for the determination of the most feasible EOR method applicable to the Garzan fields were initiated. The test results of both the TPAO's research Laboratories and HYCAL Laboratory were utilized ir this study. At the end of the feasibility studies, SSI's basic findings and recommendations for the Garzan L.elds can be summarized as follows: a) The history matched original-oil-in-place figures for the Garzan-B and C pools are 195.7 MMSTB and 105.1 MMSTB respectively. b) A significant reserves extension of the Garzan-B pool appears to exist along the south flank primarily in the eastern portion of the pool. c) Approximately 40 percent of each pool's original- oil-in-place is contained in the structurally lowest of the four geological layers. This layer is also the poorest quality rock relative to the others.  - 53 - d) Increasing the pumping capacities of existing wells and reinstating selected shut-in producers will accelerate the oil production. e) The current peripheral waterflood scheme can be cptimized by increasing fluid rates and voidage in areas where water has channelled through the reservoirs and watered out producers. f) Implementation of an immiscible Dodan gas (CO2) injection is estimated to provide higher recovery,but had equal economic benefits compared to water flooding. g) Steam and in-situ combustion processes are not feasible for Garzan. 3.2.9 The simulation of Raman reservoir was conducted with the "KAPPA" model. KAPPA is a double porosity-permeability, variable API, faulted black oil simulator. The original-oil-in- place amount was estimated by volumetric means through initializations done with a conventional black oil simulator (BETA II/F). Following the reservoir characterization, the history match was carried out in 6 segments covering large, representative and critical areas. In total, 4 three-dimensional and 2 cross-sectional area history match studies were done with KAPPA. The subsequent performance sensitivities, the "do nothing" base case and the optimized base case prediction runs were also made with KAPPA on selected representative regions of the reservoir. The reason for studying Raman in different segments was due to computer limitations. T'HA total field model consisted of over 6000 (58x26x4) grid blocks. SSI's basic findings and recommendations for the Raman field can be summarized as follows: a) The degree of the development of the fracture system shows variations in the field. b) The reservoir production mechanism is controlled by the presence of fractures and a very active steady state bottom aquifer. Aquifer is separated from the Mardin reservoir interval by a dense platformal limestone.  - 54 - c) Each well interacts with the bottom water aquifer which lies directly below it in accordance with the specific collection of factors and reservoir characteristics in that area. d) The recovery factor is lower in the northern and eastern regions than in the southern and southwestern area. e) The total original oil-in Place in Raman is 611.4 MMSTB. 594.3 MMSTB of the oil is in Mardin, and 17.1 MMSTB of the oil is in the Garzan interval. f) Base case predictions estimate a recovery increase from 43.4 MMSTB at the end of 1984 to 62.8 MMSTB by the end of 2002, indicating a recovery increase from 7.1 % to 10.3 % overalJ. (Figure 20). g) Field performance can be improved by optimizing the current operations by recompleting the wells, operating them with higher capacity pumps, and by producing the wells until the economic water cut limit of 95 % is reached. h) Operating under optimized conditions, full field recovery by 2002 can be increased to 78.3 MMSTB, meaning an increase of 34.9 MMSTB from 1984. i) In-situ combustion, steamflooding or cyclic steam stimulation are technically not viable and not recommended. j) Cyclic CO2 is not recommended due to its poor recovery. Continuous CO2 injection can not be recommended either, since its recovery is less than the water flood recovery. Fractures reduce the effectiveness of CO2 injection schemes in Raman. k! Optimized CO2 (WAG) injection, under optimistic assumptions, is not sufficiently better thar. the waterflood to justify the additional expenses of a field-wide scale project. But in conmarison to the waterflood alternative it exhibits nearly the same recovery potential, if not be' Therefore, it could,still be tried in the field the form of a pilot test.  - 55 - 1) Based on abo