In the entire Ohio-Mississippi River System there is no other point where a given amount of storage for increasing low water flow would be so valuable. Cone wango would double the minimum flow of the Ohio at Pittsburgh and for some distance below, and would increase the minimum flow through the entire length of the Ohio and lower Mississippi Rivers. Because of salt beds, cheap coal, and abundant labor, this area promises to be the chemical industry center for America, the chief limiting factor being the water supply. Conewango would conserve that to the utmost. Kinzua would waste the greater part of it.

ADVANTAGES OF CONEWANGO

The advantages of Conewango over Kinzua to which we have applied financial estimates, using the minimum figures in all cases, amount to more than $100 million, and if less than the maximum figures are used, to about $400 million. It is almost embarrassing to mention such large figures in speaking of the advantages of Conewango over Kinzua, when a fraction of that difference should be conclusive. The fact is that in the Conewango setting and natural enormous outlet channel to Lake Erie we have a most unusual situation. In probably not one reservoir building job in a thousand would one find such a combination of favorable conditions. But there it is, and these surprising values naturally follow from it. Why throw them away for the sake of not admitting

an error of 30 years ago?

It may be argued that if Kinzua is built now, in a few years when the air is cleared, Conewango can then be built. But this would mean a great loss. The irreplaceable waterlevel highway along the Allegheny would have been destroyed. The exceptionally fine recreation area along the Allegheny would have been wiped out. And since both reservoirs are not needed, most of the cost of building Kinzua would have been wasted.

THE COST OF CONEWANGO

Now as to cost estimates for constructing Conewango. Our estimate is $60 million less than that of the Tippetts firm for the same degree of protection they planned. About half that difference is due to the fact that the Tippetts firm overlooked the best of the possibilities, that is, the outlet to Lake Erie by Cattaraugus Creek. When they started their study I told them that I had only begun to look into the many possibilities of Conewango and that further exploration might disclose great opportunities for economies. I urged them to make those further studies. They did not do that, but only looked into the specific suggestions I was able to make at the time they began. I continued my inquiry, and turned up other possibilities so much superior as to make my earlier suggestions obsolete. When I asked the Tippetts engineers to look into other possibilities I had found, they replied that they did not have time to do so. They never made a general study of the situation, and never qualified themselves to pass on all the major alternatives.

Diversion down the Cattaraugus, which the Tippetts engineers did not study, reduces the cost by $20 million under the plan for diversion down Silver Creek, which they did study. A proper treatment of the Pennsylvania Railroad, which they overlooked, further reduces the cost by about $9 million. These two corrections alone would reduce the Tippetts estimate of cost for Conegango to where it is no higher than Kinzua. But there are other large errors in the Tippetts estimate, to the extent of another $30 million.

Construction conditions for Conewango are strikingly different from those for Kinzua. For instance, excavation for Kinzua highways is on precipitous mountainsides, and a large part of it is in solid rock. In contrast to this, the principal earthmoving for Conewango highways is on level bottom land, where there is no rock, but only sand, gravel, and clay. There are millions of yards of this exceptionally favorable work, which would actually cost only about a third as much per cubic yard as would the earth and solid rock on the precipitous mountainside in Kinzua. The Army Engineers estimated $1 a cubic yard for this difficult mountainside earth and rock excavation. The Tippetts engi neers took that price per yard and applied it throughout for highway and railroad earthmoving for Conewango, where the real cost would be about a third as much per cubic yard.

The Tippetts engineers took the same course concerning timber clearing. For the Kinzua job, the timber clearing is mostly on steep, rocky mountainsides,

er along riverbanks where the trees lean toward the water, and fall into it when they are cut, making it necessary to haul each one out by cable. There are more than 100 miles of such riverbanks on the Kinzua project. For the Kinzua jeb this clearing was estimated at $500 an acre. On the Conewango project the timber clearing is in large tracts of flatland, where modern clearing equipment is at its best, and where the actual cost per acre of clearing would be about a third as much for Kinzua. Yet here the Tippetts firm used the same price, $500 per acre, for the broad flatlands in Conewango as was used for the steep, mountainsides and the miles of riverbank clearing on Kinzua.

This same careless process of transferring unit costs on one job to very different conditions on another job was used in relation to earth dam construction.. That is not responsible engineering.

Also, the Tippetts engineers, in estimating highway changes for Conewango, ignored the published and accepted standards of the U.S. Bureau of PublicRoads and the New York Department of Public Works, which provide that such an improvement is responsible only for meeting the extra expense made necessary by the improvement. The Tippetts engineers planned farflung highway Changes, with standards of construction far more expensive than anything now xisting or (with the single exception of U.S. Route 17) beyond anything even emotely planned in that area, and charged the whole cost to the Conewango, project, thus improperly adding millions of dollars to their estimate.

In order that there should be the least possible reason for differences of opinon about cost estimates, we used the Tippetts estimates unchanged except here we had clear reason and data for changing them. For a large part of the Tippetts estimates they give almost no data from which their estimates can be checked. We took their estimates without change in such cases, not because we think they are right, but because they gave almost no information for checking them. We believe they are too high by a further $5 million to $10 million more of improper cost.

With the corrections in unit costs which I have mentioned, and with the change of outlet and treatment of the Pennsylvania Railroad, our estimate for Conewango to fully protect Pittsburgh and Warren from three times as large a flood as would Kinzua, and to store the same amount of water for low water increase, is somewhat less than $85 million. If the project should be enlarged to cost $105 million, which still is $8 million less than the Army Engineers estimate for Kinzua, then in addition to all these values it would also store three times as much water as Kinzua for increasing the low water flow.

THE BASIS OF THE DECISION OF THE CORPS OF ENGINEERS AGAINST CONEWANGO The Tippetts engineers, in their letter presenting their report to the Corps of Engineers, made just three points against the Conewango plan. They were: First, that the plan would cost more; second, that more people would be displaced and third, that more land would be taken.

The charge that Conewango would cost more is inaccurate.

As to persons displaced, the 6 villages in the Kinzua Reservoir would be completely buried under water, and according to the explicit statement of the Army Engineers, 90 percent of the population in the Kinzua reservoir would have to be removed to entirely new locations at a considerable distance.

In the Conewango area, on the contrary, at least 75 percent of the persons affected live in 6 villages around the margins of the marsh. The parts of these villages which might be flooded would be moved up hill a fraction of a mile without interrupting the community life. These are extremely quiet places in an economically depressed area. In one of these villages only one new housea four-room cottage has been built in the past 100 years. These old villages would he changed to live communities in one of the best recreation areas in

the region.

The Tippetts report improperly included the entire population of these villages among the persons displaced by the Conewango improvement. If these villages are omitted from the count, the displacement is far greater for Kinzua than for Conewango. Therefore, that criticism is misleading and inappropriate. Moreover, the Tippetts report failed to count approximately 400 of the 835 Indians who would be displaced by the Kinzua project. Marsh has little value. Omitting villages mentioned, the market value of the values. The Conewango

land, as estimated by the Corps of Engineers and by the Tippetts engineers, an including the State park land, is greater for Kinzua than for Conewango. Tha criticism of Conewango has no validity.

These three reasons are the only ones given by the Corps of Engineers to th public and to the Congress in announcing the decision rejecting the Conewang plan.

An engineering oversight of more than 30 years ago never has been com petently reviewed during the years since, either by the Corps of Engineers or by engineers employed by them. If not corrected, it will cause the loss of hun dreds of millions of dollars in cost and values. If I should be able to contribute in some degree to preventing that loss, and to avoid the breaking of the oldest living treaty of our country, I should feel that I had at least justified my food and lodging for my lifetime.

APPENDIX

HOW THE CORPS OF ENGINEERS CHECKED ESTIMATES FOR THE DIVERSION
PROJECT TO LAKE ERIE

When I testified before the Senate subcommittee in April 1957, my testimony and that of Barton M. Jones was given by Senator Ellender of the Senate subcommittee to the Corps of Engineers with a request that our estimates be appraised. Senator Ellender, chairman of the subcommittee, said:

"At this point I wish to state that I am going to ask Mr. Tofani to bring the suggestion of Mr. Morgan to the attention of the [Army] Engineers, and when I recall them in the next 2 weeks that they be prepared and give us an estimate of the costs and advisability of such a method to control the waters of the Allegheny.

"*** I wish to give you assurance that before this committee acts we shall go into the details of your views. * * * And I hope that the engineers give you the same treatment that you give to them. ** I have been dealing with them [the Army Engineers] 20 years now. I haven't found them wanting" (public works appropriation, 1958, pp. 2464, 2471, and 2473).

Without doubt the action of Congress has been in part based on the assurances given them by the Corps of Engineers.

The reply of the corps to the Senate committee was characterized by a lack of understanding and an inaccuracy which we can scarcely imagine coming from a great public agency. The Conewango setting makes possible several different types of design. One of these, as described by Barton Jones in his testimony, calls for a very large channel, with 160 million yards of excavation, to carry all floodwater quickly to Lake Erie below ground level, with almost no reservoir storage, and therefore requiring the taking of relatively little land— in this case 8,000 acres. It would, in fact, drain and reclaim several thousand acres of the Conewango Marsh. Another type of design called for large storage capacity in that same marsh, requiring the taking of 32,000 acres of land, but calling for the excavation of 60 million cubic yards of earth-100 million cubic yards less.

In estimating the cost for the Senate committee, the corps took the big esti. mate of 160 million cubic yards of excavation from one type of plan, and the big estimate of 32,000 acres of land from another very different design, and included the cost of both as though they were part of one design. It would be physically impossible for these two major elements to be parts of the same plan.

The land to be taken by the large channel would be mostly in the lowest part of the Conewango Marsh. The actual sales prices for the low land which would be occupied by this channel, where recorded by the Tippetts engineers, runs from $6 to $30 per acre. The estimate presented by General Person is $567 an acre. Thus, General Person's estimate of the cost of land required is more than 75 times what it should be.

It is clear that whoever made this estimate was ignorant of the fundamental conditions of the plans for which he was making an estimate, and similarly uninformed as to the nature of the land taken and as to its going value. Of course, this resulted in a totally false and inaccurate estimate of cost.

This impossible combination of elements from strikingly different plans was accompanied by an excessive estimate of unit cost and the mistaken classification of material. The Corps of Engineers estimated that the work proposed in Mr. Jones' testimony would cost $200 million as against Mr. Jones' estimate of

$80 million. The largest single item in the estimate of the corps, as estimated by General Person, was for construction of a section of large channel calling for the excavation of 77 million cubic yards of glacial sand, clay, and gravel.

Now the Tippetts engineers, who made their study a few months later, in the employ of the corps, had a large item for substantially identical material, to be excavated under less favorable circumstances. In that largest item in the project, their estimated cost, for identical material under less favorable conditions, was considerably less than half as much per cubic yard as in the estimate presented by General Person for the Corps of Engineers.

The work involved in this very large item is standard large scale excavation, with no complications, and the cost can be estimated within quite narrow limits. On such work a difference of 25 percent either way from the mean would represent a very wide range for competent bidders who are financially and otherwise equipped to do the work to advantage. The variation of more than 100 percent in the estimates for such work is beyond all reason. It does not represent responsible estimating.

For another large item of 25 million cubic yards of identical material, that in the diversion channel, the estimated cost per cubic yard given to Senator Ellender by General Person for the corps is more than twice as high per cubic yard as that estimated by the Tippetts engineers for that identical material a few months later.

In addition, the estimate which General Person presented to the Senate committee improperly classified more than 20 million cubic yards of excava tion, and estimated it at more than 5 times the cost per cubic yard estimated by the Tippetts firm for substantially identical material. On these three earth moving items alone the estimate of the Corps of Engineers, as given to the Senate committee, is about $50 million higher than it would be at the Tippetts estimate of unit cost for the same material.

The very careful and detailed analysis we had made of this cost indicated that the Tippetts estimates are somewhat high. Our estimates were made by a man whose work for 25 years has been to make detailed working estimates by which large contractors make their bids on large earth moving contracts. The estimates he made for this work were in the same careful detail with which he makes estimates for contractors' bids.

The estimate made by General Person for the Senate committee might be excused on the ground that in the short space of 2 weeks no adequate estimate could be made, and that mistakes were probable. Yet when General Person presented this estimate to the Senate committee, he testified: "We feel the studies have been completed, and that further studies would not affect our conclusion." That has been the position of the corps all along.

Is it not a serious matter that when a Senate committee specifically calls for the judgment of a great national agency as to the adequacy of an estimate it is considering, that this agency should present a statement prepared without even a rudimentary understanding of the work under consideration, and that its estimates of cost should be more than double those made by responsible engineers? Under such circumstances, what chance is there for a fair hearing to be had? This, I regret to say, is somewhat typical of the publicity which has issued from the Pittsburgh office of the corps.

Mr. MORGAN. There are some points on which the air has been cleared during the past 2 or 3 years. One of them is this: This is accepted by the Army Engineers and by their consultants and by everybody that the Conewango Reservoir which we believe is better would control and completely protect Pittsburgh and the lower cities against three times as great a flood as would Kinzua. There is no dispute about that in the facts of the case. Kinzua would protect against a pretty good sized flood, but the flood catastrophes of this country come from

The chief differences were that the length and size of the channel in the Person's item was larger, and that therefore the cost of equipment per cubic yard would be less in the item estimated by the Corps of Engineers than in the item estimated by Tippetts; and that the larger dimensions of the Persons item would be more favorable to the method of excavation proposed (by Tippetts), the use of suction dredge. The actual unit cost for the Persons item would be about 15 percent or 20 percent less than for the Tippetts item.

what are termed as unprecedented waters. Up in New England 5 years ago we had an unprecedented flood, but it occurred. According to the Army estimates it cost $1.5 billion.

Now, the Corps of Engineers is getting ready to protect against that sized flood.

In the city of Dayton the greatest flood in 150 years had been about 90,000 cubic feet a second and then a flood came along in 1913 of 250,000 cubic feet a second. It was unprecedented, and yet it came. If the additional protection were to cost a billion dollars when you can get it at, we say, less cost even than the Kinzua, to leave out for Pittsburgh that protection against a very unusual flood, that would be an extreme disaster if it should come. That is not good public policy. The fact that it would protect against three times as big a flood is what is important, and there is no question on that on any side.

Second-and I will point out here on the map as I go along the Kinzua Reservoir is this shown in red. Our program is to divert all of the waters of the Allegheny River into a depression just to the north that would hold three times as much water as Kinzua Dam would; part of that to be used for raising low waterflow and part of it for holding floodflow and then out of the upper end of the Conewango Reservoir we would cut through to the Cattaraugus Creek, which is a gorge 300 feet deep with rock walls 100 feet high. That goes down to Lake Erie and any excess flood could be turned down into Lake Erie almost without limit. In my tenure of 50 years in engineering I have known of no other case where it was so possible to turn vast quantities of water away from the place where they would be damaging without the cost of building extra reservoirs for the purpose of getting rid of it.

According to the Chief of Engineers, in a speech which he made 3 or 4 months ago, the Corps of Engineers sees the expenditure of $4 billion in the Ohio Valley for water control in the next 20 years. Another statement they made is that $2 billion of that will go for reservoirs. About one-half of that, or about $1 billion, goes for the purpose of reducing floods on the Ohio.

Now, through the opportunity of turning that extra water over to Lake Erie where it is always wanted, we can eliminate up to 2 million acre-feet and prevent floodwaters in a wet year from going down the Ohio and completely eliminate the towns along the Ohio. However, in such a flood as that the Kinzua Dam would only hold it back about 5 days and then you have to let it run on down. But, this way it is out of the way forever.

The Corps of Engineers is planning hundreds of millions of dollars-about $2 billion-for reservoirs on the Ohio which have been approved by the corps, and of those about one-half of that is for flood control on the Ohio. I am speaking in rough terms. It might be a little more or a little less. To get that much control by reservoirs would cost at least $100 or possibly $200 million. This is a rare thing to be able to turn water out of the way. We have to store water and you cannot turn it out of the way. That water would be worth not less than $100 million, but it might be worth $200 million and you get that as an extra.

This control through the Conewango Reservoir, we say, will cost about $30 million less than the recognized cost of Kinzua, but there