The invention consists in a device and in a method to control pickling processes for carbon steels, austenitic, ferritic and martensitic stainless steels, duplex steels and special alloys, in which said device automatically manages sampling of pickling baths and analysing of said samples in order to define (according to specific conductivity and potentiometric methodologies) critic process parameters and to restore desired concentrations of necessary chemicals in the pickling tanks. The invention also permits to manage pickling conditions specific for the type of steel under treatment through definition of remotely activable operative procedures automatically recalling and realising the most apt operating conditions for pickling of the specific kind of material under treatment.

In the rolling, drawing, extrusion, heat treatment of steel products (such as plates, strips, tubes, rods) oxide layers are formed on the surface thereof which must be removed both to get proper final appearance as well as passivity and anticorrosive properties for the final product, and to allow further working.

Said superficial oxide layers are usually eliminated by a chemical treatment (pickling) based on exposition of the metallic material to the action of one or more acid baths containing inorganic mineral acids (sulphuric, hydrochloric, nitric, hydrofluoric) alone or mixed with one another, at proper dilution and temperature, followed by at least one final rinsing in water.

For stainless steels, the usual pickling processes (either by immersion, spraying or turbulence) require a mixture of nitric and hydrofluoric acids; such processes entrain very serious ecological problems due to emission of the reaction byproducts (extremely toxic nitrogen oxides) into the atmosphere as well as of great quantities of nitrates into waste water.

Hence, during the recent past a number of alternative "ecological" processes have been devised characterised by the elimination of nitric acid.

Among such processes, particularly effective at the industrial scale are those utilising mixtures of sulphuric or hydrochloric acid, hydrofluoric acid and ferric ions, in which the proper concentration of such ions into the pickling bath is maintained through addition of hydrogen dioxide. Some of such processes are described in Italian patents 1,245,594 and 1.255.655 (corresponding to US-A-5 345 383) and in European patent application EP-A-0 769 575.

In the traditional pickling technotogy according to the above-mentioned patents, the management of the process usually includes an occasional control of the pickling bath through manual titration of the acidity or measure of the conductivity of the solution and of its iron content (or of total metals, through measurement of bath density); it is also possible to measure the content in hydrofluoric acid by means of a specific ion selective electrode.

Some of these techniques have been utilised in the automation of single operations in nitric acid based pickling processes of stainless steels.

US patent 4,060,717 (LECO Corp.) discloses the use of ions selective electrodes for fluorine and hydrogen ions to measure the concentration of nitric acid (or other strong acid) and of hydrofluoric acid in pickling baths containing nitric and hydrofluoric acids; the electric voltage data gathered by a control circuit are elaborated by a microprocessor to calculate the concentration of the two acids and to adjust relevant concentrations.

JP patent 55040908 (NIPPON Steel Corp.) discloses the determination of the hydrofluoric acid and of another strong acid (nitric, hydrochloric, sulphuric) through the determination with ion selective electrodes of the relevant anions after passing the solution through ion exchange membranes, in order to adjust the acids concentration.

US patent 5,286,368 (FOXBORO Corp.) measures the concentration of hydrofluoric acid in a mixture of nitric and hydrofluoric acids through the complexing ability of trivalent iron ions towards the fluorine ions, permitting to determine the concentration of the acids in the mixture.

JP patent 072944509 (KAWASAKI Steel Corp.) measures the concentrations of free hydrofluoric and nitric acids and that of iron ion in a pickling solution by measuring the concentration of iron ion by an iron salicylate complex absorptiometric method, the concentration of free hydrofluoric acid by an iron acetylacetone complex fading absorptiometric method and the total concentration of free acids by neutralising titration method, the concentration of free nitric acid being measured by subtracting the concentration of free hydrofluoric acid from the total concentration of free acids.

JP patent 081660003 (MITSUBISHI Heavy Ind. Ltd.) refers to a method for continuously measuring the iron ion concentration in a pickling solution.

The continuous automatic management of such pickling processes based on nitric acid, though better than an occasional manual or automatic control performed, for instance, a few times per day, is not essential for the process in terms of quality of treated material, because of the functional characteristics of such baths; particularly, in the pickling of stainless steels, such baths usually have high nitric acid concentrations (about 12-15%) and hydrofluoric acid concentration of about 2-5%. The high nitric acid concentration ensures at the same time both high acidity and almost constant oxidising power, making it possible to manage the process through occasional additions of chemicals. Moreover, the determination of acid concentration is sufficient to have an adequate control of the pickling ability of the bath.

On the contrary, the nitric acid free pickling systems, such as those previously cited, found the oxidising properties of the system on the measure of the ferric ions (Fe³⁺) concentration, or better on the control of the Fe³⁺/Fe²⁺ ratio.

In this case, because of the pickling reaction (1)

        2 Fe³⁺+ Fe0 → 3 Fe²⁺     (1)

in a continuous process for the production of stainless steel strips or in automatic, high productivity plants for rod pickling, the trivalent iron ions concentration, the Fe³⁺/Fe²⁺ ratio and hence the oxidising capacity of the solution tend to quickly diminish, continuously and drastically modifying the bath behaviour.

The optimal conditions must be, therefore, continuously adjusted by means of oxidising agents, such as hydrogen peroxide.

Moreover, the variation of the trivalent iron concentration indirectly influences also the concentration of the free acids present into the bath.

For instance, in a pickling system based on sulphuric acid, hydrofluoric acid and ferric salts mixtures, this influence is linked to the following preferred equilibria:

        Fe³⁺ + n F^- → FeF_n^(3-n)+

        F_e²⁺ + SO₄^2- → FeSO₄

Hence, during the oxidation/reduction reaction of the couple Fe³⁺/Fe²⁺ liberation of respectively sulphuric acid and of hydrofluoric acid will occur from relevant complex salts, thus modifying the bath composition.

A process control through occasional analytic measures, followed by large additions of chemicals to restore the best pickling conditions, causes, therefore, too ample variations of the bath parameters with adverse consequences on the product quality and on the process costs.

On the other hand, frequent manual controls and relevant composition adjustments are time consuming and costly, since this requires a large amount of personnel to ensure a satisfactory control frequency (e.g. a control per hour).

The criticity of nitric acid free pickling processes is obviously linked to the total iron amount dissolved per time unit, to the number of pickling tanks to be controlled, to the number of materials requiring different operative conditions and to the practical capability to ask for frequent manual additions of acids into the tanks.

The management of pickling processes for stainless steels such as those previously cited for continuous pickling plants of stainless steel strips or for high productivity automatic plants for rod processing, proved to be critic for the quality of the final product; it can also be non economic without the use of an automatic system for sampling, controlling and dosing of the reactants.

The control device and the method according to the present invention require the utilisation of specific skilfulness and analytical methods for a proper management of such processes.

It is an object of the present invention to provide a control device for nitric acid free pickling baths comprising means to take a sample of the bath to be analysed; means to analyse said sample in order to measure a number of parameters according to specific conductivity and potentiometric methodologies as well as to measure the redox potential value of said sample and its temperature; restoring means, apt to calculate, according to the above measured values, the quantity of correction chemicals to be added to the pickling bath in order to restore at the desired level the value of said parameters and to actuate at least a device to add into said pickling bath said quantities of correction chemicals; wherein said measured parameters are the concentrations of sulphuric acid and of hydrofluoric acid measured according to specific conductivity methodologies and of bivalent and trivalent iron ions measured according to specific potentiometric methodologies.

It is a further object of the present invention to provide a method for controlling nitric acid free pickling baths, comprising at least the following steps:

• taking a sample of a pickling bath;
• measuring according to specific conductivity methodologies the concentration of the sulphuric and the hydrofluoric acid in said sample of a pickling bath;
• measuring according to specific potentiometric methodologies the concentration of the bivalent iron ion in said sample of a pickling bath;
• measuring according to specific potentiometric methodologies the concentration of the trivalent iron ion in said sample of a pickling bath;
• measuring the redox potential of said sample of a pickling bath;
• measuring the temperature of said sample of a pickling bath;
• restoring at preset levels the values of said measured concentrations in said pickling bath by adding calculated quantity of correction chemicals to the pickling bath.

The invention will now be described with reference to a non limiting embodiment shown in the enclosed figures where:

• Fig. 1 schematically shows a plant comprising an analysis device according to the invention;
• Fig. 2 shows a simplified scheme of an analysis device according to the invention;
• Fig. 3 schematically shows the analysis vessel CA of Fig. 2, comprising a conductivity measuring system and a preferred embodiment of the rinsing means of the vessel itself and of the measure electrode;
• Fig. 4 schematically shows the analysis vessel CA of Fig. 2, comprising a potentiometric measuring system and a preferred embodiment of the rinsing means of the vessel itself and of the measure electrode.

In the enclosed Figures, corresponding elements will be identified with same reference.

Fig. 1 schematically shows a plant comprising an analysis device according to the invention, comprising:

• a plurality of pickling tanks V (V1, .... , Vn);
• an analysis device A (later on described with reference to the simplified scheme of Fig. 2) which, in the herein described embodiment, includes a couple of analysis devices (A1, A2) simultaneously working on different parameters;
• a plurality of reservoirs S (S1, S2, S3) each containing a solution at a given concentration of one of the correction chemicals (a strong mineral acid, preferably sulphuric acid, hydrofluoric acid and an oxidising agent, preferably but not necessarily hydrogen peroxide) to be added into one of the tanks V;
• a plurality of permanent recycling piping, connecting tanks V to the sampling inputs I (Fig. 2) of the analysis device A;
• a plurality of piping to feed the correction chemicals, connecting reservoirs S to tanks V;
• addition means enabling the analysis device A to control the addition into tanks V of correction chemicals contained in reservoirs S.

For simplicity, in Fig. 1 components not interesting for the present description, such as valves, pumps, actuators, filtering and rinsing means, known per se, as well as other, if any, circuitry components are omitted.

The analysis device A comprises (Fig. 2) means to pick up from a vessel V a sample of the pickling bath; means to analyse it to measure, according to specific conductivity and potentiometric methodologies, the preset parameters (the sulphuric acid, and the hydrofluoric acid concentrations, as well as the ones of trivalent and bivalent iron), the redox potential and the temperature of said diluted sample; means to calculate the amounts of correction chemicals to be sent from reservoirs S to tanks V to adjust said parameters and means to actuate the devices at the output of reservoirs S to send into the pickling bath the calculated amounts of said correction chemicals.

Since the time necessary for measuring the concentration of sulphuric and hydrofluoric acids is shorter than that necessary for the measure of iron ions concentration (just some minutes vs about 30 minutes), the analysis devices (A1, A2) are preferably divided, each one being specialised in only one of said analysis (measure of sulphuric acid and of hydrofluoric acid, respectively of iron ions concentrations).

The analysis devices (A1, A2) can be managed by a logic unit of higher level, not shown in the figures, which can be placed "in loco" or in a remote site, connected to the analysis devices (A1, A2) through bi-directional transmission means, known per se.

Alternatively, said analysis devices (A1, A2) can be of the same model and comprise the analytical means apt to measure the concentration both of the acids (sulphuric and hydrofluoric) and of the iron ions.

In such a case, the device according to the invention could also work in case of malfunction of one of the analysis devices (A1, A2).

Fig. 2 shows a simplified scheme of an analysis device A (A1, A2) of Fig. 1. comprising in combination relationship:

• a sampling module C, the sampling inputs of which I (I1, ...., In) are in sequence connected to the permanent recycling piping among the pickling tanks V (V1, ..., Vn; Fig. 1) and the analysis device A; at least a reservoir (not shown), in which the bath sample to be analysed is loaded, is provided inside the sampling module C;
• a reagent storage DR, containing the chemicals for the analyses;
• dosing means D (D1, D2) apt to draw the amounts of chemicals necessary to the analyses and to transfer the same into the analysis vessel CA, part of the dosing means D being apt to draw with low accuracy (from about 2 to about 5%) high quantities of chemicals, the remaining dosing means being apt to draw with high accuracy (about 0,1%) small quantities of chemicals; in Fig. 2 the dosing means D with low and high accuracy are respectively grouped in two different functional units (D1, D2).;
• an analysis vessel CA, containing the measure electrodes (generically named EM in Fig. 2), receiving from sampling module C the bath sample to be analysed, from dosing means D the chemicals necessary for the analysis and from a reservoir W (not shown) the water (preferably having a conductivity lesser than 100 microsiemens) necessary to dilute said sample to a desired dilution ratio; in Fig. 2 further elements (such as stirrers) present in analysis vessel CA are omitted, not being part of present invention;
• a logic unit UL, controlling and managing the analysis procedures, acquiring and elaborating the information from measure electrodes EM and actuating means to send into the pickling bath the solutions of the correction chemicals contained in the reservoirs S (Fig. 1).

In a preferred but not limiting embodiment the dosing means of functional unit D1 are peristaltic pumps with constant delivery, while the dosing means of functional unit D2 are syringes in antacid material (e.g. PES) operated by an electric stepping motor.

Again in a preferred embodiment, the analysis device A also comprises means (hereinafter described with reference to Figures 3 and 4) permitting to rinse the analysis vessel CA and the measure electrodes EM after each measure with water and after a given number of measures with a chemical solution (preferably but not necessarily 10-20% hydrochloric acid), thus permitting to keep in optimal conditions the measure electrodes EM, to have reliable analytical data, to reduce to a minimum the maintenance interventions and to highly enhance the electrodes life.

To ensure a constant quality of the final product, each type or family of materials to be pickled must be treated according to standard and characteristic parameters (hydrofluoric and sulphuric acids concentration, trivalent and bivalent iron ions concentration, ratio between trivalent and bivalent iron ions, hydrogen peroxide concentration, temperature of the sample to be analysed, and so on); in a preferred embodiment of the invention, the parameters characterising each working step as well as those concerning the operation of the analysis device A, which all permit to perform different analyses on pickling baths relating to the specific working step, are grouped into operating procedures biuniquely correlated with the material itself and stored in the logic unit UL, which are when necessary recalled according to the material to be pickled.

Preferably but not necessarily, an operating procedure comprises at least the following information:

• the order and the kind of the analyses to be performed;
• the prefixed values of the parameters for the pickling bath;
• the magnitude of the admissible deviations with respect to the prefixed values, beyond which the logic unit UL actuates said means to send into the pickling bath the solutions of the correction chemicals contained in reservoirs S,
• the dilution ratios with water in the analysis vessel CA of the pickling bath sample to be analysed.

The proper operation of the analysis device A can be advantageously checked periodically and automatically; to this end, in a preferred embodiment of the present invention a further operative autocalibration procedure is stored in the logic unit UL which activates after a given number of analyses and comprises the functional steps of drawing from a container (preferably but not necessarily located in the reagent storage DR) a fixed amount of a standard solution having a known composition, of transferring it into the analysis vessel CA, of analysing it, of comparing the obtained analytical results with the known composition and of activating alarm signals if the deviation between obtained analytical results and known concentrations is larger than a desired value.

According to an embodiment of present invention, not shown in the figures, the logic unit UL can be connected to a central operative post and/or to a logic unit of higher level, by which it can be controlled and managed; as above said, this logic unit of higher level can be placed "in situ" or be remote.

In particular, at each change of working activity, the central post operator can modify the operative procedure performed by one or more of the logic units UL, activating the one pertaining to the activity to be initiated; the operator can also recall from one or more of the logic units UL an operating procedure, modify it and have it to be performed by the logic units UL and/or inputting a new operative procedure storing it in the logic units UL.

The analytical methods, which are utilised in the analysis of the pickling baths, will now be described to better understand the described details, which are part of present invention.

This determination is based on the principle that, in an aqueous solution formed by a mixture of a weak acid such as hydrofluoric acid and of a stronger acid such as sulphuric acid, the solution conductivity is practically equivalent to the one of the strong acid at the same concentration; the method also exploits (in a stage subsequent to a first conductivity measure on a bath sample duly diluted to measure the sulphuric acid concentration) the high affinity of hydrofluoric acid for a metal cation present in the solution as a salt of known concentration. The salt anion most preferably come from a strong acid (e.g. nitric or hydrochloric acid) so that the reaction forming fluorocomplexes of the metal cation and hydrofluoric acid will generate a significant increase of conductivity due to the formation of an equivalent amount of fully dissociated strong acid, measured by a second conductivity measure.

For instance:

        n HF + Fe(NO₃)₃→ FeF_n^(3-n)+ + n HNO₃

Such conductivity increase is, therefore, proportional to the concentration of hydrofluoric acid which, after a proper calibration, can be quantitatively measured.

Such salts can be, for instance, ferric nitrate, ferric chloride, aluminium nitrate, aluminium chloride; in a preferred embodiment of the invention a solution of ferric nitrate*9H₂O is utilised, at a concentration of 750 g/l.

To ensure a sufficiently linear dependence of conductivity from the variation of acids concentration, the sample dilution must be attentively evaluated as a function of the concentration of the acids present in the bath to be analysed; as a non-limiting example, for sulphuric acid concentrations up to 200 g/l and for hydrofluoric acid concentrations up to 60 g/l, dilution ratios from 1:100 to 5:100, and preferably 4:100, are deemed to be acceptable.

Another variable essential for the obtainment of reliable results (which must be managed by the logic unit UL of the analysis device A) is the sample temperature after dilution with water; in fact, in the industry the water temperature can have considerable variations (usually between +5 and +40° C) according to the weather, to the water source and to the holding time in reservoir W.

It is apparent that a conductivity measure is greatly influenced by the temperature, and usually such a problem is overcome by means of an automatic compensation system incorporated into the measure device; in the present case, the automatic compensation can correctly adjust only the effect on the first conductivity measure (determination of the sulphuric acid concentration) but not on the second one (determination of the hydrofluoric acid concentration) performed after addition of ferric nitrate as the solution composition is changed and its dependence from the temperature is, in fact, different before and after the ferric nitrate addition.

This critic problem is solved with an analysis device A according to the invention, in which the logic unit UL takes into account the conductivity variation due to the addition of a volume v3 of the ferric nitrate solution, depending on the sample temperature.

The amount of ferric nitrate utilised during the titration must be such to ensure a full complexing of the hydrofluoric acid; in the considered system, for hydrofluoric acid concentration less than 60 g/l the ratio between the volume v3 of a solution of ferric nitrate*9H₂O at 750 g/l and the volume v1 of the bath sample must be higher than 0,5 and preferably 1.

As a non limiting example the following operating procedure is given along with relevant computations for a sample dilution of 4:100 in volume:

• filling of the analysis vessel CA, by means of dosing means D2, with a given water volume v2, having a conductivity of less than 100 microsiemens to obtain a dilution ratio of 4:100;
• picking up from the sampling module C (by means of dosing means D2) of a given volume v1 of the pickling bath sample to be analysed;
• start stirring the solution;
• first conductivity measure (L₁);
• addition of a given volume v3 = v1 of a solution of ferric nitrate*9H₂O at 750 g/l;
• stirring of the solution and measuring of its temperature T;
• second conductivity measure (L₂).

The logic unit UL acquires the data L₁, L₂, T and automatically find the concentration of the acids through the following calculations:

• sulphuric acid concentration (g/l): a · L₁² + b · L₁ -c
• hydrofluoric acid concentration (g/l): a₁ · δ² + b₁ · δ - c₁

• δ = L₂ - L₁ - φ;
• φ = c₂+ (c₃ · T);
• c₂, c₃ are constants depending on the quantity of ferric nitrate*9H₂O added to the diluted sample before the second conductivity measure.

In this example:

• a = 0,0066; b = 5,015; c = 6,98
• a₁ = 0,0120; b₁ = 2,881; c₁ = 3,81;
• c₂ = 9,632; c₃ = 0,297.

Fig. 3 shows the characteristics of the conductivity cell CC, which specific form allows to minimise the negative effects due to the high viscosity of the solution and to facilitate the rinsing of the measure platinum plaques.

Said conductivity cell CC comprises a hollow body B, in glass and having a substantially cylindrical shape, containing two blackened platinum plaques EL; at the lower and upper parts of the hollow body B there are holes (F1, F2) letting the sample to be analysed to circulate inside the hollow body B.

Preferably, the hollow body B has a diameter of about 20 mm (and anyhow comprised between about 17 and 23 mm) and a height of about 40 mm (and anyhow comprised between about 35 and 45 mm); the EL plaques dimensions are about 10 x 5 mm (and anyhow between about 8 x 12 mm and about 3 x 7mm), the distance from one another being about 15 mm (and anyhow between about 12 and 18 mm).

To avoid polarisation of electrodes EL, the measure electric circuit (not shown) connected to the conductivity cell CC must work at high frequency (between 25 and 40 kHz).

The bivalent iron determination can be made through potentiometric analysis, by potassium permanganate titration according to the classic methodology.

The operative sequence requires:

• pouring into the analysis vessel CA a given water volume v2, through the overflow pipe TP, to obtain a dilution ratio ≥ 1:50;
• picking up from the sampling module C (by means of dosing means D2) a given volume v1 of the pickling bath sample to be analysed, and addition of said sample into the analysis vessel CA;
• acidification of the diluted pickling bath sample by means of addition into the analysis vessel CA (by means of dosing means D1) of a given non-critical amount of a solution of a strong acid, e,g, a sulphuric acid solution 1:1 bw;
• potentiometric titration, having a preset final point or with an automatic search of the final point with a 0,1N potassium permanganate solution added into analysis vessel CA by means of dosing means D2;
• emptying and rinsing analysis vessel CA.

The trivalent iron is measured by iodometric titration, utilising however some specific attention to permit the use of an automatic device and the obtention of reliable and reproducible results.

Said determination requires the following operating sequence:

• pouring into the analysis vessel CA a given water volume v2, through the overflow pipe TP, to obtain a dilution ratio ≥ 1:50;
• picking up from the sampling module C (by means of dosing means D2) a given volume v1 of the pickling bath sample to be analysed, and addition of said sample into analysis vessel CA;
• start of stirring;
• addition into analysis vessel CA (by means of dosing means D1) of a given non-critical volume of a lanthanum nitrate solution having a known concentration;
• waiting for 30 s without stirring;
• addition into analysis vessel CA (by means of dosing means D1) of a given non-critical volume of a hydrochloric acid solution at 1:1 vol;
• addition into analysis vessel CA (by means of dosing means D1) of a given non-critical volume of a potassium iodide solution, at a concentration for instance of 1 kg/l;
• waiting for 5 min without stirring;
• start of solution stirring;
• potentiometric titration with 0,1N sodium thiosulphate (added by means of dosing means D2) of the iodine liberated by the reaction of trivalent iron with potassium iodide;
• emptying and water rinsing of analysis vessel CA.

For this automatic analysis, a most prominent aspect is the use of lanthanum nitrate; in fact, the addition of a salt including a cation able to complex the fluorine ion linked to the iron ion is essential for the quantitative analysis of the ferric ion through iodometric analysis.

This analysis can be manually performed utilising a solution of calcium chloride; however it was proved that calcium chloride cannot be utilised for the automatic titration of trivalent iron, because of the subsequent precipitation of calcium fluoride and of calcium sulphate, which tend to continuously foul the electrodes in analysis vessel CA, giving rise to significant errors and complex upkeeping. On the contrary, it was found that lanthanum salts can quantitatively release the ferric ion, generating powdery and non-sticking lanthanum fluoride precipitates, thus permitting the automatic management of the process with high reliability and very limited upkeeping.

This same result can also be achieved by adding to the system a complexing agent for the iron ion, which however can quantitatively release it during the subsequent reaction with potassium iodide; complexing agents such as EDTA can be fit for this purpose.

The potentiometric system, schematically illustrated in Fig. 4, comprises a measure electrode E (inert to the working environment) immersed in analysis vessel CA and a reference electrode R (preferably in glass, of the type Ag/AgCl) positioned outside said analysis vessel CA and in contact with the solution under measurement through a saline bridge, comprising an electrolyte (contained in a tank SR) which is made to continuously pass through a porous septum SP placed at an extremity of a small plastic tube T.

The continuous passage of the electrolyte through the septum SP is intended to consent the electric continuity; to avoid the contact between septum SP and the hydrofluoric acid of the pickling bath and to continuously renovate the electrolyte.

In a preferred embodiment, the measure electrode E is made from a body in antacid material bearing at one extremity a platinum plaque P, one of whose surfaces, mirror finished, faces downwards, thus preventing the salts deriving from the reaction products to be deposed on the measuring face of plaque P, fouling it.

Advantageously, to the electrolyte (preferably 3M potassium chloride) can be added a 10% solution of glycerine (of another compatible product having a viscosity at 20°C comprised between 1,15 and 1,45 centipoise, inert with respect to the working environment and functionally equivalent) to enhance the viscosity and reduce the flow speed, thus permitting a better autonomy of the potentiometric system for a given volume of tank SR.

The determination of the free hydrogen peroxide in nitric acid free pickling processes such as the ones here described, is necessary in the treatment of ferritic and martensitic steels for the control of finishing/passivating baths generally utilised as the last operation before final rinsing; usually said baths comprise sulphuric acid (20-60 g/l), hydrogen peroxide (3-10 g/l) and sometimes hydrofluoric acid.

The analytical methodology and the operative sequence utilised for the determination of the hydrogen peroxide are the same utilised for the determination of bivalent iron in the pickling baths.

The device according to the invention measure, before the determination of bivalent iron, the solution redox potential on the diluted pickling bath sample utilising the potentiometric system already described; the thus obtained value is very near (± 20 mV) to the redox potential measured in the bath before its dilution.

The obtained value is compared with a range of values (usually comprised between 200 and 550 mV) stored into the logic unit UL to be utilised as a first signal of the correct operation of the system: if the measured value is outside of said range, the logic unit UL of the analysis device A stops the analysis procedure and sends an alarm. The calibration of the potentiometric system is made at a given frequency (say, once per week) by redox potential measure on a standard solution of known potential (usually 468 mV).

As already said, the logic unit UL of an analysis device 1 according to present invention, after measuring the desired parameters on the pickling bath sample under analysis, calculates the amount of each of the solutions at known concentration of the correction chemicals (sulphuric acid, hydrofluoric acid and oxidising agent) contained in reservoirs S, said chemicals being opportunely added to the pickling bath to restore the desired composition values and actuates addition means (such as, for instance, dosing pumps or electrovalves) at the output of reservoirs S to send into the pickling bath said calculated amounts of the correction chemicals.

Being known the plant characteristics (volume of tank V, delivery of each adding means, preset concentration values for said correction chemicals, concentration of said chemicals, and so on) to have the correct amount of correction chemicals added to the pickling bath, the logic unit UL must just calculate the actuating period of said addition means.

Studies and experiments of this Applicants did show that, to bring back to the desired values the concentrations in the pickling bath of sulphuric acid, of hydrofluoric acid, of trivalent iron ion and of the oxidising reagent, the logic unit UL must actuate each of the addition means regulating the addition into the pickling bath of the sulphuric acid, hydrofluoric acid and oxidising reagent solutions, for a period of time s (in seconds) given by the following expression:$$\mathrm{s}=\mathrm{K}\cdot ({\mathrm{v}}_0-{\mathrm{v}}_{\mathrm{m}})\cdot {\mathrm{v}}_{\mathrm{b}}/\mathrm{p}$$
in which:

• s = actuating time (seconds);
• K = factor inversely proportional to the concentration of the correction chemicals (I/g);
• v₀ = given concentration for the specific corrective chemical (g/I);
• v_m = concentration of said specific corrective chemical resulting from the analysis (g/l);
• V_b = volume of tank V;
• p = delivery of the addition means (l/s).

To bring back to the desired value the ratio R between the concentration of trivalent and bivalent iron ions, the logic unit UL calculates the actuating period s1 (in seconds) of the addition means sending into the pickling bath the oxidising reagent solution, by:

• calculating B₁ = A · R, in which A is the concentration (g/l) of the bivalent iron ion resulting from the titration with permanganate, R is the desired ratio between the concentration of, respectively, the trivalent and the bivalent iron ions, and B₁ is the theoretical concentration of the trivalent iron ion;
• comparing B₁ with the measured concentration B of the trivalent iron ion (g/l);
• if B ≥ B₁ (the measured concentration of the trivalent iron ions is greater than that of the bivalent ones) the logic unit UL does not act;
• if B < B₁ (the trivalent iron ions concentration is less than the measured one) the logic unit UL calculates the actuating period s1 of the addition means regulating the addition to the pickling bath of the oxidant reagent solution, by means of the formula$$\mathrm{s}1=\mathrm{K}\cdot {\mathrm{K}}_1\cdot \mathrm{C}/\mathrm{p}$$

• s₁ = actuating period (s);
• K factor inversely proportional to the concentration of the correction chemical (i/g),
• K₁ = factor proportional to the tank volume V (I);
• C = (B₁-B)/R = amount of bivalent iron ion to be oxidised to restore the desired value for iron ion concentration (g/l);
• p = delivery of the addition means (l/s).

Alternatively the bath can be managed in function of the ratio R between trivalent iron and bivalent iron according to the following calculation:

• Calculation of the total iron T = A + B

• Calculation R = B/A
• Compare R (present ratio) with R1 (pre-fixed ratio)
• If R > R1 the logic unit UL does not make any addition of oxidizing product
• If R < R1 the logic unit UL calculates the actuating period s1 (in seconds) of the addition means regulating the addition the pickling bath of the oxidizing product solution according to the following formula $$\mathrm{s}1=\mathrm{K}\cdot {\mathrm{K}}_1\cdot \mathrm{C}/\mathrm{p}$$

• C = A-[(A+B)/(R₁+1)] = amount of bivalent iron to oxidize to restore the present ratio R to the prefixed value R₁
• s1 = actuating period (s)
• K = coefficient, inversely proportional to the volume of the tank V (I)
• P = delivery of the addition means (I/s).

Fig. 3 schematically shows an exploded view of the analysis vessel CA of Fig. 2, comprising a measure system of conductivity type and a preferred embodiment of the rinsing means of analysis vessel CA and of measure cell CC.

In Fig. 3 it is possible to see:

• the conductivity measure cell CC used for conductivity measure;
• the analysis vessel CA;
• the overflow TP, mobile, the position of which (controlled by the logic unit UL) consents to set the liquid level in the analysis vessel CA, and to empty the same vessel;
• rinsing means (F, U) controlled by the logic unit UL, enabling the rinsing of analysis vessel CA and of the conductivity measure cell CC.

Fig. 4 schematically shows an exploded view of the analysis vessel CA of Fig. 2, comprising a potentiometric measure system as well as a preferred embodiment, similar to the one in Fig. 3, of the rinsing means of analysis vessel CA and of the measure electrodes.

In Fig. 4 can be seen:

• the potentiometric system, comprising the measure electrode E, the reference electrode R, positioned outside of the analysis vessel CA, and the saline bridge which in turn comprises an electrolyte contained in tank SR, continuously passing through a porous septum SP placed at one extremity of a small plastic tube T;
• the analysis vessel CA;
• the mobile overflow TP, the position of which (controlled by the logic unit UL) consents to set the liquid level in the analysis vessel CA, and to empty the same analysis vessel;
• rinsing means (F, U) controlled by the logic unit UL, enabling the rinsing of analysis vessel CA, of the electrode E extremity and of the porous septum SP.

In the preferred embodiment shown in Figures 3 and 4, such rinsing means comprise a plurality of slits F placed along the upper edge of the analysis vessel CA and a nozzle U apt to rinse with a water spray the extremity of the measure electrode E and the porous septum SP, respectively the conductivity measure cell CC; in Figures 3 and 4, can also be seen the lid CP for the analysis vessel CA and means MS supporting the electrode E, the small tube T of the potentiometric system, the conductivity measure cell CC and the small tubes (not explicitly indicated in Figures 3 and 4) connecting the dosing means D (D1, D2) with the analysis vessel CA; lid CP and supporting means MS will not be described, as known per se and anyhow not pertaining to present invention.

Preferably, the analysis vessel CA, the measure electrode E and the porous septum SP (respectively the analysis vessel CA and the conductivity measure cell CC) are water rinsed after each analysis and washed with a chemical solution after a given number of analyses.

To rinse said components with water after each analysis the logic unit UL performs in sequence the following steps:

• fully emptying analysis vessel CA;
• pouring in said analysis vessel CA a large amount of water through slits F;
• filling with water analysis vessel CA up to have the tip of electrode E and the porous septum SP, respectively the conductivity measure cell CC immersed;
• emptying analysis vessel CA;
• further rinsing the tip of electrode E and the porous septum SP, respectively the conductivity measure cell CC by spraying on them some water through nozzle U;
• emptying analysis vessel CA and preparing it for the subsequent analysis.

To wash after a given number of analyses with a chemical solution (preferably 10-20% hydrochloric acid) the analysis vessel CA, the tip of electrode E and the porous septum SP (respectively the analysis vessel CA and the conductivity measure cell CC), the logic unit UL fills with water analysis vessel CA through slits F up to have the tip of electrode E and the porous septum, respectively the conductivity measure cell CC immersed, picks up from a tank (preferably but non necessarily placed within the reagent's storage DR) an amount of product (preferably hydrochloric acid) necessary for said chemical washing and send it into analysis vessel CA; after a given period of time the logic unit UL empties analysis vessel CA and rinse it with water, to eliminate any trace of the chemical solution.

Moreover, when not working, analysis vessel CA is filled with water through slits F and nozzle U, to avoid any fouling and/or damaging of the electrode E tip, of the porous septum SP, and of the conductivity measure cell CC.

It is possible for an expert to modify and improve, as suggested by ordinary experience and by the natural technical evolution, the device for the control of pickling baths according to present description, still remaining within the scope of present invention.