|
APPENDIX 5COST OF VHF STATION IN CO. DONEGAL.F81/2/667 Clerk to Committee, Committee of Public Accounts, Leinster House, Dublin 2. I refer to your letter of 19 November, 1986 in which you request a comparison of the costs relating to the building of a coastal VHF station in Co. Donegal with the cost of similar type buildings constructed in Co. Kerry in the years 1981 to 1984. I attach for the information of the Committee a detailed cost comparison between the VHF station at Crockalough Co. Donegal, built by Department of Communications directly, and the three VHF stations built by this Office under the Valentia contract, which has been prepared by our Quantity Surveying Section. It will be noted from the details in Appendix B to the cost comparison that the total expenditure for the three stations built by this Office is put at £399,866 whereas a figure of £464,511 is mentioned in the Report of the Comptroller and Auditor General. The latter figure was the best estimate available at the time and while the final account for the total contract i.e. works to the Coast Radio Station at Valentia and the VHF stations, has yet to be agreed with the contractor it would appear that the cost of the VHF stations will be substantially below the estimated figure. Yours sincerely, Paschal Scanlan, Chairman. 24 March, 1987. VHF Stations Cork/Kerry — Crockalough Station, Co. Donegal.Cost ComparisonIntroduction1.0 We refer to your request of the 19th November, 1986 for cost comparison between the V.H.F. station at Crockalough, Co. Donegal built by the Department of Communications directly and the three V.H.F. stations built by the O.P.W. under the Valentia contract. 1.1 We have prepared the following reply based on our records of the Valentia contract and the information supplied by the Department of Communications on the Crockalough contract. Report of the Comptroller and Auditor General (C & AG)2.0 In the report of C&AG on the 1984 Appropriation Accounts, Vote 41—Communications, figures in relation to the cost of the two V.H.F. contracts are quoted. We wish to advise that both amounts have been revised in our report. 2.1 In the case of Crockalough we had deducted the non-building costs. This gives a net cost for building works of £25,057.80 as set out in Appendix A. 2.2 In the case of the total for Valentia the updated figures give a net cost for building work of £399,866 as set out in Appendix B. Cost Information — Generally3.0 There are basically three levels of cost comparisons which are determined by the level of cost information available. 3.1 In order to give a comprehensive cost comparison a priced Bill of Quantities is essential. This enables the quantity surveyor to compare the quantities, quality, and unit rates of the projects under consideration. 3.2 An elemental cost breakdown of the tenders would permit the identification of areas of cost difference. 3.3 The superficial method is where the overall cost is divided by the floor area. This method provides no indication for the reasons behind the cost difference. Cost Information — Crockalough4.0 The cost information available on Crockalough generally falls into the category of 3.3, with elemental breakdown on some items and a few unit rates. It should be noted that this level of information is not unusual on a project of this scale. The preparation of a Bill of Quantities would not be justified on a project of less than £100,000 in 1980. Quantities and Unit Rates5.0 The Bill of Quantities referred to in 3.1 collectively describe the quality and quantities of works to be undertaken. 5.1 The quality of the work is set out in the specification. It is accepted practice that where the brief to the design team is silent on the specification the standards required are left to the discretion of the architect. 5.2 The unit rates are the cost at which a contractor is prepared to undertake a single unit of a particular piece of work e.g. place concrete in foundations — unit rate £75.00 per cubic metre of concrete placed. The unit rate is made up of the cost of labour, materials and plant used in placing the concrete. Form of Contract6.0 The Valentia contract was placed on the basis of the O.P.W. form of contract. This form permits the contract sum to be adjusted due to variations, increased costs etc. and permits stage payments. 6.1 The Crockalough form was a fixed price lump sum contract. The O.P.W. uses a similar form for certain minor works. 6.2 Since March 1982 all major building contracts, whether placed through the O.P.W. or directly by a Department must be under the G.D.L.A. ’82 form of contract. The essential provisions of the O.P.W. form are incorporated into the G.D.L.A. ’82 form. 6.3 The Crockalough form of contract provides no relief to the contractor from risks which the client would normally be expected to carry e.g. substructural conditions. It would also provide no relief in the event of additional costs due to delays in either acceptance of the tender or delays in construction work outside of the control of the contractor e.g. the appointment of nominated sub-contractors. This must be considered in the light of construction costs increasing at a rate of 22% per annum in this period. Construction Industry Federation policy, both then and now, is to oppose fixed price contracts, mainly for the foregoing reasons. 6.4 The differences in the form of contract have other implications which are considered under the headings Quantities 12.0, Increased Costs 13.0, Unit Rates 14.0. 6.5 We would normally expect higher tenders to result from a fixed price form of contract as a contractor has to price the risk factor arising from 6.3, 12.0, and 13.0. Tendering Procedure7.0 The Crockalough tender was obtained by open tendering. This method has much to commend it, in particular it produces very competitive tenders. However the lowest tender is not the sole criterion for placing a contract. Considerations such as the contractors organisation, resources and financial standing are equally important and have led to the more frequent use of pre-selection, selected tendering and negotiated contracts. 7.1 In the case of the Valentia contract the extension to the existing station was let on the basis of open tendering while the V.H.F. (excluding Electrical and Mechanical) stations were based on a quotation and negotiated unit rates. Evaluation of Tenders — Generally8.0 Where we are retained on projects without a Bill of Quantities there are a number of criteria which we apply in the evaluation of tenders. 8.1 Cost Limits: Where these apply the main consideration would be the percentage below the limit which the contractor is likely to be able to complete the project without becoming bankrupt. If there was doubt we would obtain from the contractor the tender build up. This would enable us to check that the quantities and unit rates were adequate. If these items were acceptable we would allow the market forces to prevail. 8.2 Estimate: In the case where we have prepared an estimate we would first update the estimate from the original date of preparation to the tender date. This is accepted practice and does not produce a hypothetical figure but one on which a valid comparison of the estimate and tender can be prepared. As in 8.1 above we would have to satisfy ourselves regarding differences between our estimate and the tender before acceptance. 8.3 No Estimate: Where we are asked to evalute a tender where we have not prepared an estimate we prepare one at current rates for the purpose of comparison. This is the approach that was used on the parts of the Valentia contract that we were asked to evaluate. Evaluation of Tenders — Crockalough9.0 The Crockalough V.H.F. tender was accepted without reference to a cost limit or an estimate. Under these circumstances it is necessary to further consider the factors which may result in Crockalough being unsuitable as establishing a cost limit for V.H.F. stations generally or providing a suitable mean for the purposes of estimating or evaluating other tenders. Range of Tenders — Crockalough10.0 In the case of the building works there were three tenders and the percentage difference was, second lowest tender 39% above lowest and third tender 76% above lowest. Such large percentage differences are not unique but we would expect a 1-5% range normally. The percentage difference in this case would have warranted further examination. It is the experience of this office that the greatest deviation between cost limit/estimate and tender occur in the border counties. 10.1 The work in connection with the access road and hardstanding were tendered for separately and attracted four tenders. The percentage differences were, second lowest tender 56% above lowest, third lowest tender 79% above lowest and fourth tender 114% above lowest. The same comments as in 10.0 apply. 10.2 The electricial work had a 3.5% difference between the lowest and second lowest. Contractors Organisation11.0 The acceptance by the Crockalough contractors of a fixed price lump sum contract indicates that their main area of operation was domestic type construction. Such organisations would rarely employ the services of an estimator to prepare a tender or evaluate their price in the light of tenders submitted by other contractors. 11.1 The contractor on the Valentia project was undertaking the V.H.F. stations as an extra on a contract for the erection of a complex structure. They possessed an adequate organisation for this work including the services of an estimator. Quantities12.0 The fixed price form of contract offers no relief to the contractor in the event of additional quantities of work being necessary e.g. additional excavation due to lower bearing strata, soft spots etc. and additional work to rectify these items i.e. additional fill, rising walls etc. 12.1 On the Valentia contract the contractor was entitled to an architects instruction if such difficulties were encountered and consequently payment for the additional work involved. Similarly items marked ‘Provisional’ were subject to remeasurement of the actual work involved. Increased Costs13.0 The O.P.W. form of contract has provision for increased costs, the Crockalough one did not. This resulted in the time scale increasing the final cost of the Valentia V.H.F’s while Crockalough (excluding variations) remained unchanged. 13.1 The factors which caused the later starting dates for Ballyhearney and Allihies were outside of the scope of the contract. 13.2 The mast bases were requested after the completion of Kilkeaveragh and Ballyhearney and this contributed to the high cost of increased costs incurred on site works. 13.3 In the example selected for cost comparison i.e. Kilkeaveragh, increased costs have been shown both separately and distributed to demonstrate the impact of this item. (Appendix C). Unit Rates14.0 The lack of unit rates on the Crockalough project is the main factor why a quantified explanation of the differences in unit cost cannot be given. 14.1 As stated in 5.2 unit rates are made up of the cost of labour, material and plant to perform a single unit of work. In addition the cost of the contractors overheads and insurances are frequently included. This is the case in Crockalough and Valentia. With one exception V.A.T. would also be deemed to be included in Crockalough. Comparison of Unit Rates15.0 There was only one unit rate which lended itself with direct comparison. This was the unit rate for placing concrete in foundations and in the case of Crockalough appeared in the context of an extra due to additional concrete for foundations to masts. The unit rate for Crockalough was £72.00 (V.A.T not included) the basic unit rate used in Valentia was £75.00. A typical norm for this unit rate in December 1981 would be £48.00. This difference demonstrates the considerable cost implications of building in a remote location. However no useful cost comparison can be deduced from a single unit rate. 15.1 There are a number of unit rates which appear on the tender in connection with part of the road works. These appear in the report of the Inspector in Letterkenny, dated 13/10/’82. It is not clear whether materials are to be purchased or obtained from the mountain side as only costs of labour and plant are shown. Even allowing for the exclusion of material costs these rates are not economical. 15.2 There are four unit costs shown. We are unable to comment on the one for a roller as the description is too vague. The other three are as follows: 15.3 J.C.B. and driver for 5 days.........£280.000 This gives a unit rate of operating of £7.00 per hour (based on a 40 hour week). An economical rate for this operation would be £12.08 per hour. 15.4 Two tractors and trailers for 5 days ....... £500.000 This gives a unit cost of operating of £6.25 per hour (based on a 40 hour week). An economical rate for this operation would be £9.23 per hour. 15.5 Four labourers for 5 days .......£500.00. This gives a cost of employing labour of £3.13 per hour (based on a 40 hour week). The minimum ‘all in cost’ of employing labour at the date in question was £4.96 per hour. The ‘all in cost’ is calculated by adding to the agreed rate the cost to the employer of holidays, Pay Related Social Insurance, Employers Liability insurance etc. 15.6 It would appear that the unit costs for plant are below a rate that would generally be considered economical. This may be explained in part by the exceptionally low ‘all in cost’ of employing labour. 15.7 The main components in the “all in cost’ of labour are governed by various clauses in the O.P.W. form of contract. 15.8 The Fair Wages clause requires that current rates of wages agreed are paid. 15.9 The insurance clauses require that the contractor shall maintain Employer’s Liability and Public Liability Insurance. 15.10 The Wage and Price Variation clause requires that where for example increases in P.R.S.I. are claimed that the claim is substantiated by evidence of payment. 15.11 There is therefore little scope for contractors operating under the O.P.W. form to avoid incurring the full ‘all in costs’ of employing labour. 15.12 As the unit rates in 15.3 — 15.5 only appear in the context of one section of the site works at Crockalough it would be incorrect to imply that they applied throughout. However as labour accounts in broad terms for 50% of the cost of a building contract, any scope for not paying normal rates would have a significant impact on tender levels. 15.13 In addition to the insurance requirements mentioned in 15.9 the O.P.W. form requires contractors to maintain “all risks” insurance for the full value of the work. While the cost of this insurance depends on a number of factors it can be expected to add up to 0.5% to unit rates. Cost Comparison16.0 The cost comparison is based on the breakdown of figures set out in Appendix A,B,C. The explanation for the unit cost variation has been approached on the basis of the final cost of the main elements. This enables a distinction to be drawn between elements which should have a cost relationship, elements which are common to the units but are not cost comparable and elements which are not common to all or unique to one unit. Cost Comparison — Building17.0 The four V.H.F. stations have identical specifications. The only difference was that a precast concrete roof was used in Valentia while an in situ concrete roof was used in Crockalough. As this variation on the Valentia contract had no cost implications it can be ignored for purposes of cost comparison. The basic buildings differ only in terms of floor area with Crockalough and Kilkeaveragh both being 32 sq. m., Ballyhearney and Allihies both 58 sq. m.. 17.1 The floor area factor coupled with a closer completion date suggest that a more realistic cost comparison would arise between Crockalough and Kilkeaveragh, with Ballyhearney and Allihies adjusted on a pro rata basis. (Appendix B & C.) 17.2 The completion date in each case was selected on the basis of representing a date by which the bulk of the cost incurred had been paid. 17.3 An analysis of the two costs on a superficial basis (building cost — floor area) gives a cost of £247.00/sq. m. for Crockalough and £441/sq. m. for Kilkeaveragh, a cost difference of 79%. As outlined in 4.0 in the absence of further details on Crockalough no specific reasons behind the cost difference can be deducted. 17.4 In the absence of specific reasons for the cost difference we propose to provide broadly based reasons by reference to a similar building type which provides a greater numerical basis for analysis. 17.5 Since 1964 the O.P.W. has placed contracts for approximately 1,200 Rural Automatic Telephone Exchanges (R.A.X) buildings. These buildings are standardised within categories and thereby provide a suitable basis for comparison. They are frequently placed on the basis of a fixed price contract similar to the one in Crockalough. The majority are placed on the basis of open tendering. The category selected is similar in specification and scale to the V.H.F. stations. The “B.D.” R.A.X. has a gross floor area of 37 sq. m.. The one important difference is that R.A.X’s are usually located in towns, on flat sites and adjacent to services. In the case of the “B.D.” type a total of 19 tenders were available for comparison in the period May 1980 — November 1981. 17.6 There were three ‘B.D.’ tenders in the Donegal region in the relevant period. The cost per square metre, which has been adjusted to July, 1982, is given after the location. Burtonport £246 sq. m. Muff £234 sq. m. Fintown £227 sq. m. This would give an average of £236 sq. m. The Crockalough cost was £246 sq. m. However the R.A.X. cost must be adjusted to take into account location. In 15.0 dealing with unit rates an example for concrete was given at £48 cu.m. while Crockalough was £72 cu.m. As concrete forms a large component of the building its cost difference provides a suitable weighting for adjusting the cost for location. The average of £236 adjusted by this factor would be £354. This would give a total cost of £11,328 which is in line with the figure of £11,239 for the second lowest tender, when it is adjusted for Prime Cost and Provisional Sums. 17.7 We suggest that there were two major tendering errors in the lowest contractors bid. The first was an error in rates i.e. the cost would only cover a normal location, and secondly, the percentage reduction below cost in order to secure the contract. The first point is borne out by reference to a typical cost for similar work in the area and the second by the 39% difference between the lowest and second lowest tender. The market conditions in January 1981 were buoyant so it was not necessary to reduce margins by this amount in order to secure the contract. It was also inadvisable to reduce margins for the cost considerations outlined in 6.3 12.0 and 13.0. 17.8 It is not sufficient to cite the fact that the contract was completed for this amount as proof to the opposite. This was a contract of low capital value and short duration. Should the contractor have incurred a loss, this would not be evident until after completion and would probably not be of a scale to cause bankruptcy. 17.9 The error in tender strategy would account for only 50% of the cost difference of 79%. 17.10 As stated in 10.0 the greatest deviation between cost limit/estimate and tender occurs in the border counties. Allowing for the vagaries of the tendering process the contention would appear to hold good in the case of R.A.X. stations. A typical tender for the west Cork/Kerry region is based on the following cost per sq.m., adjusted to July 1982. Annascuail £355 sq.m. Inch £336 sq.m. Ballydavid £279 sq.m. Killinardin £296 sq.m. This would give an average of £317 sq.m. for the west Cork/Kerry region. A cost of £236 sq.m. for R.A.X’s in Donegal produces a percentage regional variation of 34%. This regional variation would be more than adequate to account for the outstanding cost difference of 29%. 17.11 It must be stressed that regional variation is not a factor that is always present. In the case of the ‘B.D.’ range an adjusted tender for the border counties at Glaslough, County Monaghan was £408 sq.m. while in west Cork an adjusted tender for Ross Carberry was £169 sq.m. and Castletownbere £437 sq.m. Other factors of a transient nature such as a number of large projects or a major development may also produce a variation. This was a feature of the Cork/Kerry region in this period due to development in Cork harbour and Moneypoint, Co. Clare. Cost Comparison — Variations on Building18.0 The arguments in support of the ‘Building’ Cost Comparison would also be appropriate to the cost difference in variations. Cost Comparison — Electrical19.0 The amount of £3,929 for each of the Valentia stations is approximate but is more or less correct. 19.1 Unlike the rest of the work on the Valentia V.H.F. stations the cost of the electrical work was established on the basis of open tendering. Tenders were obtained by the O.P.W. and included work to a proposed V.H.F. station at Geokaun Hill. This station was later omitted but as the electrical work was identical the tender figure was used for the later V.H.F. stations. Increased costs apply and the provision of an earth mat and lightning protection were extras. 19.2 It is not possible to provide an explanation for the cost variation in this item between Donegal and Kerry other than that both reflect market conditions as both were obtained by open tendering. Cost Comparison — Rock, Reflector Band20.0 These items did not occur on the Crockalough project and should be omitted from the total for the Valentia V.H.F.’s for the purposes of unit cost comparison. Cost Comparison — Roads and Hardstanding21.0 At Crockalough this work was largely undertaken by a different contractor to the one responsible for the main building. 21.1 Independent cost information on works of a similar specification was not available for the purposes of cost comparison. As none of the Valentia sites were sufficiently similar to Crockalough to permit a direct comparison we have based our cost comparison for this element on unit rates. In the absence of such information for Crockalough an approximate composite unit rate per sq.m. of road has been calculated. This rate is based on a total cost for Roads and Hardstanding at Crockalough of £11,900, less cattle grid £755 = £11,145 net cost. A total area has been calculated at 1,875 sq. m. (road 550 × 3 = 1650) + (hardstanding 15 × 15 = 225). This gives a composite unit rate of £5.94 sq. m. A similar exercise at Kilkeaveragh produces a composite unit rate of £22.98 sq. m. (£20,013 + V.A.T. @ 3% = £20,613) (road 55 × 6 =(330) + (hardstanding 27 × 21 = 567). 21.2 The main reasons for the cost difference are considered under various headings as follows: 21.3 Specification: The specification for Crockalough permitted the use of selected excavated material to form the surface of the road. In Valentia imported fill (hardcore) was specified. The material content accounts for £8.72 of the Valentia composite unit rate which when deducted reduces the cost to £14.26 sq. m. 21.4 Quantities: The composite unit rate and the operating unit rate 15.3 would require adjustment if there were errors in the quantities assumed by the contractor. This does not appear to be a consideration in this case as a pre-tender briefing on site was given to the contractors at Crockalough. 21.5 Components of the unit rate: The remaining composite unit rate broadly consists of plant and labour i.e. operating costs. In 15.3 reference was made to an economical rate of operation. If this factor is applied to the Valentia rate it would reduce it to £5.99 sq. m. As stated in 15.6 the low cost of ‘all in cost’ of employing labour may have influenced the rate. This factor coupled with the regional variation referred to in 17.10 would appear to adequately resolve the difference in the composite unit rate for roads and hardstandings. Cost Comparison — Roads and Hardstandings, Variations22.0 The variation at Kilkeaveragh was due to difficulties with the septic tank location. Cost Comparison — Fencing23.0 Fences were not considered necessary at Crockalough and the site was delineated by marker posts. The cost of this work was combined with a quotation for road works and has been included in the amount for roads and hardstanding. As the cost of this work at Crockalough should be insignificant the cost of fences of the Valentia project should be excluded from the unit cost comparison. Cost Comparison — Mast and Fittings24.0 The masts supplied under the Valentia contract have no parallel at Crockalough and should be excluded from the unit cost comparison. Cost Comparison — Mast Bases25.0 As stated in 24.0 the masts fitted under the Valentia contract have no parallel at Crockalough. This is reflected in the size of the bases used. The forces to be transferred from the masts to the bases is specified by the manufacturer and these in turn govern the type and size of the foundations required. The scale is reflected in the volume of concrete required, Crockalough 13 cu.m., Ballyhearney 33 cu.m, Kilkeaveragh 87 cu.m. and Allihies 85 cu.m.. Ballyhearney was a low lying site which accounts for the variation in size from the two other Valentia examples. For the purpose of unit cost comparison this item should be excluded. Cost Comparison — Adjustment of Unit Rates26.0 This item arose due to a claim by the Valentia contractor for an adjustment of the agreed rates due to uneconomical working. An adjustment of 10% was eventually agreed. 26.1 In the case of Kilkeaveragh and Ballyhearney this adjustment applied to the builders work in connection with the masts bases and was to cover the additional cost of undertaking this work out of sequence i.e. after practical completion of the V.H.F. stations. 26.2 In the case of Allihies the rates were adjusted on all the builders work. This was to cover the additional cost of uneconomical working in this remote and inaccessible location. This item has no parallel in Crockalough. For the purpose of unit cost comparison it should not apply to Kilkeaveragh or Ballyhearney as the masts bases are being omitted but it would apply to the relevant sections of Allihies. Cost Comparison — Transport Costs27.0 This item arose only on the Allihies V.H.F. It relates to the additional costs involved in having supplies of materials delivered to the site. The costs arose because it was not possible for full loads from suppliers to proceed past the junction of the Castletownbere/ Allihies road and the access road to the site due to the latters unusually steep gradient. 27.1 The items which the contractor was compensated for were blocks, hardcore and concrete. Some £15,400 of this item relates to concrete for the masts bases and should be excluded from the unit cost comparison. Cost Comparison — Increased Costs28.0 The reasons for the cost incurred under this heading have been considered in 13.0 In addition to the amount shown the figure of £3,929 for electrical work would contain a provision for increased costs. Cost Comparison — Valued Added Tax29.0 The rate of V.A.T. increased from 3% to 5% from the 1st March, 1983. 29.1 In the cost comparison between Kilkeaveragh and Crockalough the 3% rate applied to the relevant parts under consideration. For the other stations the increase added £2,050 to the unit cost at Ballyhearney and £3,872 to the unit cost of Allihies. Unit Cost Variation — Valentia30.0 The unit cost variation on the Valentia contract is summarised as follows: 30.1 Building: The cost difference is due to the increase in the floor area from 32 sq. m. to 58 sq.m. 30.2 Roads and Hardstandings: These costs differences reflect the different site conditions encountered. 30.3 Fencing: There was little cost difference where this item was required. 30.4 Masts and Fittings: No cost difference where this item was supplied under the contract. 30.5 Mast Bases: Cost difference proportional to size of base required. 30.6 Transport Costs: Applies to one site only See 27.0 for details. 30.7 Increased Cost: This item reflects the time scale involved. 30.8 Additional V.A.T.: Due to change in rate. Summary — Generally31.0 The reason for the unit cost variation is due to the cost of the main elements. Where these elements are directly comparable the main cost variation has been identified. Where these elements are not cost comparable areas of difference are identified and eliminated and the remaining areas compared. In the case of elements not common to all these elements are identified and treated separately. Summary — Building32.0 The cost comparison of the building element first considers the cost variation between the Donegal and Kerry V.H.F.’s 17.1 — 17.10. This variation is reconciled on the basis of a tender error in the Donegal contract 17.6 and a regional variation 17.9. Cost variation within the Valentia contract was due to an increase in the floor area 30.0. Summary — Roads and Hardstanding33.0 This element was not directly comparable between Donegal and Kerry due to the specification 21.3 Items which were comparable proved difficult to reconcile due to a lack of information and only a global explanation based on arguments similar to the ‘building’ element have been forwarded 21.4 — 21.5. Variations within the Valentia contract were due to the individual site conditions encountered 30.2. Summary — Increased Costs34.0 The progressive impact of this element is evident from its contribution to the higher unit costs of the later Valentia V.H.F.’s. Summary — Other Elements35.0 The remaining elements were either not significant, not cost comparable or confined to the Valentia V.H.F.’s. Conclusion36.0 The unit cost variation is largely attributable to site works. The cost of site works hinges on the scale and specification of the works which in turn is largely determined by the location of the site. When adjustment is made for the particular site requirements the unit cost variation is within the parameters created by the vagaries of costing in the construction industry. Donal Lenihan, Quantity surveying Section. 8th December, 1986. APPENDIX ‘A’CROCKALOUGHEXPENDITURE BY DEPARTMENT OF COMMUNICATIONS
APPENDIX ‘B’VALENTIA V.H.F.’s
APPENDIX ‘C’UNIT COST COMPARISON
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||