dissenting.
Although use of the replacement cost approach is valid in determining an assessment for this special-purpose brewery plant, unreliability and uncertainty in the assessment process will be the likely result of approving the vague and overly general elements which the county appraisers here used in implementing the replacement cost approach.
Precisely because this brewery does have “unique functional and structural characteristics,” as the trial court stated, the county appraisers should have looked closely at the real estate elements of this plant and, in order to deduct nonassessable elements, should have scrutinized carefully the machinery and equipment in this plant. Instead, the record is clear that the county appraisers looked elsewhere.
Of most concern is the fact that they relied upon plant cost figures derived chiefly from annual per-barrel production ratings. The county began with a “brewing industry” standard estimating total plant cost to be $50 per barrel of *209annual plant capacity; the county then lowered that figure to $48 per barrel of annual plant capacity. However, because that construction cost unit includes the cost of machinery and equipment, the county then proceeded upon the assumption that two-thirds of the $48 is attributable to machinery and equipment, in order to arrive at a $16 per-barrel capacity for real estate structures. Multiplying that amount by the estimated annual capacity of the facility in question, the county arrived at a gross replacement figure of $56,000,000.
Even though the county thereafter followed the replacement cost approach of Reichard-Coulston, Inc. v. Revenue Appeals Board of Northampton County, 102 Pa. Commonwealth Ct. 226, 517 A.2d 1372 (1986), appeal denied, 517 Pa. 611, 536 A.2d 1335 (1987), by subtracting a figure for depreciation and adding the estimated value of the land, the final result necessarily can be no less unreliable than the starting figure of $48 or $50 per barrel of annual plant capacity.
For that crucial starting figure, the county appraisers went to the brewing industry, unfortunately, rather than to real estate appraisal sources. In an attempt to support such an approach, the county appraisers considered breweries in Michigan and Ohio not shown to be comparable to the Pennsylvania facility in question. Construction cost estimates came, second-hand, from a brewmaster and a plant engineer.
The county also made reference to the Marshall Valuation Service Manual for the $50 per barrel figure, but the record shows that source acknowledges itself as providing only “very rough guides____” On the basis of “discussions with other persons associated with the brewing industry,” the county appraiser witness considered construction costs to be “in a range from approximately $45 to $60 per barrel of annual production capacity.”
Precisely because this property has “unique functional and structural characteristics,” (as this court’s* opinion points out) reliance upon such general relationships between *210productive capacity and construction costs, at plants elsewhere, cannot be a sufficient approach.
Even if we assume that those wide-ranging amounts could provide a valid representation of total plant cost, the county appraisers never did analyze the value of the specific machinery and equipment at the property in question, in order to arrive at an accurate amount to deduct from the total.
Merely attributing two-thirds of total plant cost per production unit to machinery and equipment, in order to deduct them, is obviously an extremely inaccurate way to proceed. The productive capacity of a plant necessarily will vary greatly, depending upon the relative efficiency of the specific machinery and equipment employed. Because well-designed modern machinery necessarily will achieve greater productivity than machinery of a poor design, the flat attribution of two-thirds of plant cost to machinery in all cases is necessarily an unspecific rule of thumb.
Although this court certainly has no power to override the findings of a trial judge on a credibility basis, the county’s valuation evidence here lacks just not credibility, but competence. Even if every word of the county witnesses is believed, their method is not competent, because it has, at best, only loose relevance to the property sought to be appraised.
This court should vacate the decision appealed from and should remand the case to the trial court for redetermination of a tax value based upon the record, but excluding the county’s appraisal evidence resting upon its overgeneralized relationships between productive capacity and plant costs.